Citation
Application of neural networks in finite impulse response filters

Material Information

Title:
Application of neural networks in finite impulse response filters
Creator:
Rezaie, Hamid Reza
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
47 [137] leaves : ; 28 cm

Thesis/Dissertation Information

Degree:
Master's ( Master of Science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Electrical Engineering, CU Denver
Degree Disciplines:
Electrical Engineering
Computer Science
Committee Chair:
Wall, Edward T.
Committee Co-Chair:
Anderson, Marvin

Subjects

Subjects / Keywords:
Neural circuitry -- Mathematical models ( lcsh )
Neural circuitry -- Mathematical models ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 46-47).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Department of Computer Science and Engineering
Statement of Responsibility:
by Hamid Reza Rezaie.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
19782917 ( OCLC )
ocm19782917

Full Text
APPLICATION OF NEURAL NETWORKS IN
FINITE IMPULSE RESPONSE FILTERS
by
Hamid Reza Rezaie
B.S., University of Colorado at Denver, 1982
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Department of Electrical Engineering and Computer Science
1988


This thesis for the Master of Science degree by
Hamid R. Rezaie
has been approved for the
Department of
Electrical Engineering and Computer Science
by
Edward T. Wall
Marvin Anderson
Date
2-z /?s-?r


Rezaie, Hamid Reza (M.S., Electrical Engineering)
Application of Neural Networks in FIR Filters
Thesis directed by Professor Edward T. Wall
In this report an application of neural networks in finite
impulse response (FIR) filters has been investigated. In
chapter I the idea of neural networks is introduced and
their mathematical backgrounds for a single_unit,
single_layer and multi_layer Parallel Distributed Processing
(PDP) models have been described. Then a learning algorithm
called, the back propagation rule is chosen for training the
PDP models. In chapter II an introduction to FIR filters
and their design using windows is included and then an
equivalent PDP model, named FIR Neural Filter, has been
introduced. In chapter III, this model has been implemented
using a program, called "bp", which is written in C__language
and simulates the back propagation algorithm. Finally a
sinusoidal and an unit step response of this FIR Neural
Filter is obtained and their results are compared to the
desired responses.
The form and content of this abstract are approved. I recommend
its publication.
Signed
Edward T. Wall


CONTENTS.
CHAPTER
I. NEURAL NETWORKS............................... 1
Introduction................................ 1
Parallel Distributed Processing (PDP)....... 3
Mathematical Representation of PDP models... 6
Single Unit Representation................ 6
Two-layer representation.................. 7
Multi-layer representation................ 9
Learning.................................. 9
The Back Propagation Algorithm............... 12
II. FINITE IMPULSE RESPONSE NEURAL FILTERS......... 15
FIR Filters.................................. 15
Design of FIR Filters using Windows.......... 17
FIR Neural Filters........................... 18
n-Sample Delay Network..................... 21
Impulse Response Network................... 25
III. SIMULATION RESULTS............................. 27
Implementation............................... 27


V
CONTENTS (continued)
Sinusoidal Response.......................... 31
Unit Step Response............................ 36
System Identification and Modeling........... 43
REFERENCES............................................. 4 6
APPENDIX
A. Command Set used in the Simulations.......... 4 8
B. Simulation files used for Sinusoidal, Unit
Step Inputs and System Identification and
Modeling..................................... 50


FIGURES
Figure
1. A typical Multi_layer PDP Model....................4
2. A Single Processing Unit in PDP Models........... 6
3. A two_layer PDP Model Mapping....................8
4. A three_layer PDP Model Mapping...................10
5. A Direct Form realization of FIR Filters..........16
6. An Adaptive FIR Neural Filter.....................20
7. A Sequential Neural Network.......................22
8. A one_delay or Z Network......................... 23
9. The n-sample Delay Network with Circular Shift
property...........................................24
10. The Impulse Response Network..................... 25
11. Logistic Activation Function Characteristics..... 28
12. Impulse Response of an ideal Linear-Phase Lowpass
FIR Filter.........................................30
13. n-sample Delay Network Outputs for Sinusoidal
Input Patterns.................................. 37
14. Sinusoidal Response of Impulse Response Network
with Desire Outputs of n-sample Delay Network as
inputs............................................ 38


VII
FIGURES (continued)
Figure
15. Sinusoidal Response of Impulse Response Network
with Actual Outputs of n-sample Delay Network as
inputs............................................. 39
16. n-sample Delay Network Outputs for Unit Step
input Patterns.........;.......................... 4 0
17. Unit Step Response of Impulse Response Network
with Actual Outputs of n-sample Delay Network as
inputs............................................. 41
10. Unit Step Response of Impulse Response Network
with Desire Outputs of n-sample Delay Network as
inputs............................................. 42
19. System Identification Model........................ 4 3
20. The Plant Network.................................. 4 4
21. The Plant Transfer Function........................ 4 4
22. Output Responses of the Plant and the Neural
Network........................................ 45


CHAPTER I
NEURAL NETWORKS
Introduction
What is intelligence? How can a human brain learn
and analyze multitask operations? A computer can perform
on the order of millions of double-precision multiplication
in a second, but it does not exhibit natural intelligence.
Symbol-processing machines have failed to provide a useful
framework for capturing the simple insight required by the
interactive nature of processing.
The power of thinking of a human brain has always
been a fascinating subject to scientists and in the last
two decades there has been extensive research and effort in
this area. It is believed that the interaction between
brain cells, called neurons, is the major power of learning
capability. For example a simple task such as picking up a
pawn from the chess board involves many parallel processes
in the brain. The eyes will relay the pattern information
received from the pawn to the brain, then the arm starts
moving toward the chess board with an angle information
received from the brain and finally the fingers will pick
up that specific pawn from the chess board. These tasks


2
seem to require mechanisms in which each aspect of the
information in the situation can act on other aspects,
simultaneously influencing other aspects and being
influenced by them.
It has always been an ultimate goal to be able to
build a machine that can learn and think by itself. Some
may think that the answer is "software", but this does not
seem to be the case. An other approach is to try and build
a machine that is structured and behaves as close as
possible to a brain as we know. This is the back-bone of
"Neural Networks".
A class of models called Parallel Distributed
Processing (PDP) have adopted this idea that information
processing takes place through the interaction of a large
number of simple processing elements called "units", each
sending excitatory and inhibitory signals to other units,
'while receiving the same type of signals from other units.
The neural network is at its infancy and there is
still so much to be learned and experimented. From the
study of brain physiology and simulation results of PDP
models, new techniques and ideas for future PDP models are
being discovered. Once these and other questions are
resolved, there will result more powerful models and
hopefully a breakthrough in science.


3
Parallel Distributed_____processing__Maflaflj
The parallel distributed processing or PDP models
consist of a number of units interconnected with each other
by a set of connections called "weights". These models are
constructed to resemble the brain structure which consists
of a large number of highly interconnected elements
(neuron) which send simple excitatory and inhibitory
messages to each other and update their excitation on the
basis of these simple messages. Therefore the PDP models
are much closer to a brain structure than any other
information-processing models.
The PDP models detect and extract common features
from the patterns describing the objects that belong to the
same pattern class. If the pattern presented does not
belong to any existing class, a new class is created.
Decisions concerning the class membership of the patterns
are established by the use of decision functions [1].
Decision functions are the equations of a line, plane or a
hyper-plane which separates the boundaries of different
pattern classes in an n-dimensional space.
In PDP models, the pattern themselves are not
stored, instead the "connection strengths" between units
that allow these patterns to be recreated are stored.
Learning is accomplished through finding the right
connection strengths so that the right pattern of


4
activation will be produced under the right circumstances.
This is one of the most important features of PDP models
since it captures the interdependencies between activations
that is exposed to in the course of processing, and
produces an information processing mechanism that could
learn by tuning its connections to other units.
The general structure of a PDP model is shown in
figure 1. This network consists of three layers. The
Figure 1: A typical multi-layer PDP model


5
first layer, named input layer, has n input units which are
connected to the k units of the. second layer called
intermediate or hidden layer. The hidden layer units are
also connected to the m units of the third layer known as
the output layer.
In two layer networks, if the input represented is
such that the similarity structure of the input and output
patterns are very different, then'the network will be
unable to perform the necessary mapping. This problem can
be solved by addition of an extra layer between the input
and output layer, namely the hidden layer. This layer can
provide internal representation and therefore perform the
necessary mapping.
Another issue in PDP models is determination of
number of the hidden units and connections required for a
specific task. Although there are some ideas as how to
approach this problem, there is no real guidelines since
each problem has its own requirements. In general, if we
have the right connections from the input units to a enough
set of hidden units, we can perform any mapping from input
to output through these hidden units.
For an example consider the Perceptron model by
Rosenblat in which Minsky and Papert [2] have shown that
this model can not solve the exclusive-or problem since
those patterns that overlap least are supposed to generate


6
identical output values. This network only with addition
of a hidden layer which can create its own internal
representation of the input can solve the exclusive-or
problem.
Mathematical Representation of PDP Models
In order to get more insight into the PDP models,
it is helpful to understand its mathematical representation
and how the units in the model interact with each other.
First a single unit is described and then the idea is
extended to more complex and multi-layer networks.
Single Unit Representation
A single unit in PDP models can be described as a
unit which receives n inputs from the n units of the lower
layer as shown in figure 2. The output unit has a scalar
Figure 2: A Single Processing Unit in PDP Models


7
value called "activation" denoted byEoutand each input unit
activation is denoted by Eini. The input activation can be
represented by an n-dimensional column vector Ain where the
ith component is the activation of the ith input unit.
Each input and output unit is connected by a link
which has a scalar value called "weight". The weight
values between inputs and output unit can also be
represented by an n-dimensional row vector W. The
activation of the output unit is defined as the inner
product of the weight vector and the input activation
vector.
Eout = W Ain = [w 1 W 2
w
]
^Ein 1^
Ein2
VEinn J
(1)
Therefore the output activation is the summation of
the products of each input activation unit by its weight
value to the output unit. From equation (1) it is clear
that an output unit can actually indicate how close the
input vector is to the weight vector.
Two-Laver__Representation
From the single unit representation we know that
the output activation is the inner product of the weight
vector and the input activation vector. This idea can be


8
extended so that m output units are receiving n inputs from
n input units as shown in figure 3. The output activation
of each unit can be written as:
aouti = Wi. Ain
(2)
Figure 3: A two_layer PDP model mapping. Adapted from
James L. McClelland and David Rumelhart, Parallel
Distributed Processing(The MIT Press, 1987).
Lets define Aout and w as the vectors whose components are
Hout and Wi, respectively. Then the matrix representation
of this model is:
Aout = W Ain or
^aout i ^ ^Wll W12 .Win'' ^ainl ^
aout2 W21 W 2 2 . W 2n ain2
^aoutm J \W m 1 W m2 . W m n/ ^ainn J
(3)


9
Thus each output unit can compute the closeness of its
weight vector to the input vector, and the closer the two
vectors are the larger output activation is obtained.
Multi-Laver__Representation
A multi-layer PDP model consists of three or more
layers cascaded together as shown in figure 4. The output
of the first layer is inputted to the second layer, the
output of the second layer is inputted to the third layer
and so on. The input vector Ain is first mapped to the
hidden layer through multiplication by weight matrix Wi and
then the result is mapped to the output unit by matrix
multiplication Wx This network can be shown by only one
mapping from input to output as:
Aout = W2.Wl.Ain = W Ain where W = W2.W1 (4)
Therefore, the input vector Ain and output vector Aout are
related to each other through the weight matrix W.
Signing
Learning in PDP models involve finding the right
connections between the units so that the desired output
can be created when the input pattern is presented to the
network. It has been shown that a multi-layer PDP model is
represented by the equation (4) where each row of the
matrix w is the weight vector associated with each upper


10
Figure 4: A three_layer PDP model mapping. Adapted from
James L. McClelland and David Rumelhart, Parallel
Distributed Processing(The MIT Press, 1987).
unit. Therefore a system that can associates a particular
output vector Aout to an input vector Ain with a matrix w
has the learning capability. If the ith weight vector is
given by the outer product of Aout and Ain as :
Wi = 3.outi Ain (5)
then any desire component of Aout can be generated through
the appropriate weight vectors when presented with Ain. In
order to find a system with this property, the matrix w in


equation (4) has to be written in a way whose rows are in
the form of equation (5), i.e.:
W = Aout. A in (6)
where
AviA /Wll W12 . Win \ ^3outl ^
W21 W21 W22 . W2n 3out2
w = (JVmn J \Wml Wm2 . . Wmn J Aout = ^3outm J
and A in = ini 3 i n 2 . . 3 i n n]
As can be seen from this equation, any Wij is equal
to the product of 3outi and 3inj, which is the product of the
ith output and jth input activations. These quantities are
calculated and available by local processes and therefore a
learning scheme for finding a matrix w which will
associates any particular pair of input and output vectors
has been established.
Minsky and Papert have shown [3] that even the .
perceptron convergence procedure or the modified version of
Windrow and Hoff (delta rule) can find a learning rule for
the problems that can be solved without any hidden units,
presently there is no equally powerful rule for learning
with hidden units.


12
There has been a great deal of studying in
determination of how to create useful hidden units which
makes more abstract problems solvable. One is represented
by competitive learning [4] in which simple "unsupervised"
learning rules are employed so that useful hidden units are
developed. The other is to simply assume an internal
representation, on some priori information, which seems
reasonable as McClelland and Rumelhart [5] adopted in
interactive activation model of word perception. The third
approach is to attempt to develop a learning procedure
capable of learning an internal representation adequate for
performing the task at hand such as the one used in
Boltzman Machines [6] .
For deterministic units, yet another algorithm
called "Back Propagation Rule" [7] is also available, which
involves only local computations.
The Back propagation rule
The basic idea of the back propagation method of
learning is to combine a non-linear perceptron-like system
capable of decision with the objective error function of
Least-Mean-Square [8] (LMS) and gradient descent
procedures.
In this algorithm a set of input and output pattern
pairs are presented and the network will compute its own


13
output vector. Then a comparison between this output
vector and the desired output vector is made and the error
or difference between these vectors is calculated. This
error will propagate backward into the previous layer units
causing changes in the weights in order to minimize the
error. There will be no learning if the error is zero.
The rule for changing weights in the standard delta
rule, with no hidden units, after an input/output pair
presentation is:
ApWji = T| ( tpj Opj ) ipi = T| 8pj ipi (7)
where tpj and Opj are the jth components of the target and the
actual output due to pattern p, respectively. 9pj is the
output error and ipi is the ith element of the input pattern,
ApWji is the value of weight to be changed from ith to jth
unit after presentation of pattern p.
Let
Ep=^-^(tpj-Opj)2 be the
measure of error on
J
input/output pattern p and let the total measure of error
be E = Xep. In- order to implement gradient descent in E the
weight changes should be according to:
ApWji = 0Ep / 3wji = T| 3pj Opi
The error for the output unit is:
(8)
dpj = (tpj Opj) f j(netpj)
(9)


14
where netpj is the net total output netpj = ^Wji Opi and,/j(netpi) i s
i
the derivative of the semilinear activation function. The
error for a non-output unit is calculated as:
dpi = f j(netpj) dpk Wkj (10)
K
Therefore the back propagation rule can be summarized as
follow: The output value Opj for each unit is calculated
when the input pattern is propagated forward in the
network. Then the error signal3pjfor each output unit is
evaluated which is simply the difference between the actual
and desire output value times the derivative of the
activation function. This error is propagated backward to
each unit in the hidden layer and the appropriate weight
values from hidden to output units are changed. Then the
error signal for each unit in the hidden layer is
calculated and propagated to the lower layer. This process
continues till the total error is below an error criteria
threshold, set for obtaining the desire output values.


CHAPTER II
FINITE IMPULSE RESPONSE NEURAL FILTERS
FIR FILTERS
Discrete-time systems described by linear constant
coefficients have the general transfer function of the
form:
H(Z) =
Y(Z)
X(Z)
M
X bk z'k
k=0_________
(1- X atz-k)
k=l
(11)
The input and output of such a system is related through
the difference equation:
N M
y(n) = X ^ y(n'k> + X bk X(n-k) (12)
k=l k=0
This system can be implemented by a computational
algorithm in which the delay values of the output and input
are multiplied by coeff icients ak and bk, respectively and the
summation of these products will result in the system
output. The system function of a causal system with finite
impulse response, as shown in figure 5, is of the form:
H(Z) = X h(n) Z"
TP=0
(13)


16
Therefore if the duration of the impulse response
is N samples long, thenH(Z) is a polynomial in Z of degree
X(n)
-1
Z
----b~
-1 -1 -1
z X. z X z X
? 9 J T
h(0) ' 'h(1) ' 'h(2) ' rh(N-2)
-1 > - ^ l > L
h(N-1)
---->-
y(n)
Figure 5: A Direct form realization of FIR filters. Adopted
from Alan V. Oppenheim/Ronald W. Schafer, Digital Signal
Processing (Prentice Hall Inc., 1975).
N-l. ThusH(Z) has N-l poles at Z = 0 and N-l zeros
anywhere on the finite Z-plane. The frequency response of
this system is:
N-l
H(eiw) = £ h(n) e-J1 (14)
n=0
so the design of a FIR filter may be accomplished by
finding its impulse response coefficients. Thus the direct
form realization of a discrete-time FIR filter can be
written by the convolution sum relationship:
N-l
y(n) = £ h(k) X(n-k) (15)
k=0
where h(k) and X(n-k) are the impulse response and input
sample delays, respectively, andy(n) is the filter output.


17
Design of FIR filters_______Ji.g.inq. .Windbag
One approach in designing a FIR filter [9] is to
truncate an infinite-duration impulse response sequence and
obtain a finite length impulse response. Let Hd(eiw) be the
ideal desired frequency response defined as:
eo
HdCei") = hd(n) e-Jwn (16)
where hd(n) is the corresponding impulse response sequence,
i. e:
71
Hd(n) = ^ J Hd(eiw) eO^O dw (17)
-n
and the way to truncate the infinite impulse response is to
let h(n) be:
h(n) =hd(n) , 0 = 0 , otherwise (is)
or in another word define h(n)as the product of the desired
impulse response and a finite duration "window" W(n) such
that:
h(n) = hd(n) W(n) (19)
where W(n) is a rectangular window with the property:
W(n) =1 , 0 = 0 , otherwise (20)
and if the impulse response satisfies the condition
h(n) =h(N-l-n) then the filter has a linear phase property.


18
For example in order to design a causal linear-
phase ideal low pass filter, the desire frequency response
is defined as:
-jw(N-l)
HdCeJw) = e 2 ; Iwl < Wc
= 0 ; = otherwise (2i)
whereWcis the cutoff frequency, and the corresponding
impulse response is:
wc
hd(n) =
J><^ d
-Wfc
Sin [wc
[wc^
(22)
Then to create a finite-duration linear-phase causal filter
of length N, we define h(n) as in equation (19) .
FIR Neural Filters
FIR neural filters are the discrete-time FIR
filters which are implemented by neural networks. These
networks have the capability of learning the functions
required in order to generate filter output y(n). As
mentioned before, the direct form realization of discrete-
time FIR filters is given by:
N-l
y(n) = £ h(k) X(n-k) = h(0) X(n) + h(l) X(n-l) +.+ h(N-l) X[n-(N-1)]
k=0
so in order to implement this filter the neural system must
be able to generate N-l delays of the input X(n), then
multiply each delay by the corresponding impulse response


19
h(k) and finally calculate the summation of all these
products to produce filter outputy(n).
The FIR neural networks consist of two major
networks, a "n-sample delay network" and an "impulse
response network" as shown in figure 6. The n-sample delay
network is trained to receive the n input samples serially
and generate the input and its n-1 delay samples. This
network will shift the samples one unit to the right
everytime an input sample is presented. The sample shifted
from the right most unit will be circulated to the left
most unit, i.e. n-sample delay network will implement a
circular shift function.
The impulse response network requires only N-1
samples of the input for every computation of an output
valuey(n). Therefore, from n outputs of n-sample delay
network only the left most N-1 output units are fed into
the impulse response network. This network will learn the
N impulse response samplesh (0), h (1)..,h(N-l) internally
and multiplies these samples by inputs X(n), X (n-1),.....
X[n-(N-1)] and outputs y (n) by the summation of these
products. Therefore, this network is adaptive in a sense
that it will learn the N impulse response samples of any
given input/output pairs.
The first step before operation of the system is to
train each network by presenting input/output pairs and let


20
the network learn the desire mapping from input to output
through the back propagation rule algorithm. Once each
network is trained, the X(n) samples can be presented to
t
y(n)
IMPULSE RESPONSE NETWORK
i t i i i k
so S1
-V k *
,, i i a
X X
3 3

S2
k i i i
S S i
i N-3 1 N-2 I
JTJfA-
A A A
N-1
x
3
i
ro
x
3

z
+
n -SAMPLE DELAY NETWORK
X(n)
Figure 6: An Adaptive FIR Neural Filter
the n-sample delay network which outputs the input X(n)
and its n-1 delay samples. Then N-1 output samples of this
network is fed into the impulse response network


21
sequentially and the output y(n) is generated by use of N
impulse responses stored, internally. This process is
repeated for every sample X(n) which is presented as an
input to the filter.
At first input sample presentation switch SO is
closed andy(O) is generated by the impulse response
network. Switch SI is closed by next input sample to
producey(l) and.switch S2 is closed at the next sample and
so on. This process continues till N-l switches are
closed. From this point on all the switches will remain
closed and the input pattern is presented till all output
samples are generated.
n-Sample Delay Network
Let X(t) be a continuous signal and its samples be
represented by the values of X(n). The basic function of
n-sample delay network is to generate the input X (n) and
its n-l delay samples. In order to implement this function
with a neural network, it is necessary to introduce some
kind of dynamics into the system so that it can remember
the current as well as the previous states of the system.
This objective can be accomplished by use of sequential
neural networks [10].
The general form of sequential neural networks
consist of a set of input units and another.set of input


22
units called "current_state" units. These units are
connected to some hidden units which in turn feed into a
set of output units called "next_state" units, as shown in
figure 7. From this architecture a simple one_delay or Z-1
next state units
Figure 7: A Sequential Neural Network. Adopted from M. I.
Jordan, Attractor Dynamics and Parallelism in a
Connectionist Sequential Machine, proceedings of the Eight
Annual Conference of the Cognitive Science Society, 1986.
(Hillsdale, NJ: Lawrence Erlbaum Associates)


23
network can be implemented as shown in figure 8. First the
input X(n) is presented and the current_state input is set
to zero, then X(n) is mapped to itself and it is also fed
to the current_state unit so that it can be mapped to the
next_state unit at the next sample presentation.
X(n) X(n-1)
Figure 8: A one-delay or Z-1 Network
Therefore, everytime a sample X(n) is presented the
current_state is mapped to the next_state unit and the
current input X(n) becomes the current_state of the
network. This network will learn to generate X(n) and its
one delay sample within one percent accuracy in about 60 to
100 iterations. A more general form of this network which


24
has the circular shift property beside generation of n
delay samples of input X(n) is shown in figure 9. This
network consists of three layers, an input layer, a hidden
layer, and an output layer. The input layer has an input
X(n) X(n-1) X(n-2) X(0)
Figure 9: The n-sample Delay Network with Circular Shift Property
and n-1 current_state units and the hidden layer consist
of n-1 units which are connected to a set of n-1 next_state
or output units. The network outputs are shifted by one to
the right everytime an input sample is presented and the
outputs shifted from the last unit will be circulated to
the first unit. This network will present n delays of
input samples again if the input patterns are presented
repeatedly.


25
Tppalafi__RegRgaafi_Network
The impulse response network is a three_layer
feedforward network, in which each unit can only receive
inputs from the preceding units. This network consists of
N input units, N hidden units and N output units where N is
the number of impulse response samples. The structure of
this network is shown in figure 10. The function of this
network can be described by equation (15).
Y(n)
Figure 10: The Impulse Response Network
In order to train the network to accomplish this
task, an input/output pair which consists of X(0) and N-l


26
zeros as input is presented, which corresponds to having
switch SO closed and all other switches open. Learning is
turned on and after the required iterations the network
will learn the value of h(0) by adjusting its weight
values. The next input/output pair, i.e. X(l) and X(0) and
N-2 zeros as inputs and h (0) X (1)+h (1) X (0) as output, is then
presented. Again after the network has learned this
pattern, the value ofh(l) is also stored internally and
summation of the products h (0) X (1) andh(l)X(0) is calculated
as the output unit. This process continues till the
network has learned all N samples of the impulse response.
Now the learning is turned off and the network will always
outputsy(n) based on its inputs and the impulse response
samples stored internally. Therefore, this network is an
adaptive network in a sense that it will adjust its weight
values in response to any input/output pairs. Once the
'network has learned the impulse response samples, then no
more learning is required and output y(n) is always
generated depending upon the input presented.


CHAPTER III
SIMULATION RESULTS
Implementation
The implementation of back propagation algorithm
requires an activation function which its derivative exist.
In all the simulations a "logistic" activation function of
the form:
/j(netpj) = Opj = - where netpj = Wji Opi + 0j (23)
(i+e-netpj) r
has been used, where 0j is the bias or threshold of unit j.
The characteristics of this activation function is sho'wn in
figure 11. Therefore, the error signal 8pj for an output
unit can be formulated from equation (9) as:
dpj = (tpj Opj) Opj (1 Opj) (24)
and the error for a hidden unit is derived from equation
(10) :
9pj = Opj (1 Opj) ^ 9pk Wkj (25)
K
Also in order to have a rapid learning cycle, equation (7)
can be modified to include a momentum term a which will


28
Figure 11: Logistic activation function Characteristics
increase the learning rate T| without leading to oscillation.
Awji (n+1) = T| 0pj Opi + Ot Awji (n) (26)
where n is the presentation number.
The back propagation algorithm with the above equations is
implemented using a program called "bp" written in
C_language [11]. This program assumes that all networks
are feedforward only and consists of a main program with
many subroutines which perform the necessary computations
required in this algorithm. All the simulations are run on
a IBM PC AT with a DOS version 2.0 or higher.
The program also requires a set of input files
prior to actual implementation. A network specification
file which describes the network architecture, i.e. number
of input, hidden and output units and the connection
between them. A template file that defines the simulation


29
output result format on the CRT screen, a start file which
initializes the parameters in the program, a pattern file
that contains the input/output pair patterns, and a weight
file which contains the weight and bias of the units in the
network.
The main program reads the information in these
files and calls the appropriate subroutines for
computations needed in various parts of the algorithm.
This program also contains an interactive command set that
can be accessed during the simulation process. For more
details refer to appendix A.
In order to implement the n-sample delay network,
this-program has to be modified so that the current_state
units can receive inputs from higher order units, i.e.
output units. This is done by setting the activation of
the current_state unit, with the desire activation. If the
activation of a current_state unit i is initially set to
zero and its input in the pattern file is set to -j, where
j is the output unit, then the activation of current_state
i at time t+1 is equal to (Ej + ai), where Rj and Ri are the
activations of the output unit j and current_state unit i
at time t The constant mJLl, which is between 0 and 1,
determines the effect of the past activation of unit i on
its current activation value. Therefore, ifmjlis chosen to
be a small value, then the current state activation of unit


30
i.is always the delayed activation of output unit j by one
cycle.
An ideal low pass FIR filter of length N=ll with
twenty-one samples of sinusoidal and unit step inputs have
been simulated and the corresponding results are listed for
each input. The impulse response of this filter with a
rectangular window and cutoff frequency wc = Tt/2 can be
generated from equation (22) as is shown in figure 12.
Figure 12: Impulse Response of an ideal Linear-Phase
Lowpass FIR Filter
The n-delay sample network has one input unit,
twenty current_state units, twenty_one hidden units and


31
twenty_one output units as shown in figure 9. The impulse
response network is made of eleven input units, eleven
hidden units and one output unit as shown in figure 10.
Sinusoidal___Sg.ap.Qh9Q
In the first implementation of the low pass FIR
filter, a sinusoidal signal is selected as an input. Since
the activation value of a logistic activation function is
between 0 and 1, the sinusoidal input is shifted so that
its values are also between 0 and 1, i.e. it is defined as
X(t) =0.5 Sint + 0.5. This signal is sampled and its
first twenty_one samples are selected as the input
patterns. These patterns together with their desire output
patterns are presented to the n-sample delay network so
that it would learn the necessary mapping by adjusting
network weights between the units.
The necessary files required prior to running the
bp program are included in appendix B. The "n-sample.net"
file defines the network architecture and consists of three
sections. In the definitions section, the total number of
units and number of input/output units are specified. In
the network section, the connections between units are
defined by using a macro "%r". For example the statement
"%r 23 1 1 2" means that unit 23 and no other consecutive
unit, since 23 is followed by 1, will receive connections


32
from two units starting with unit 1 and the next one i.e.
unit 2. The letter "r" states that these weight values are
random and modifiable by the program during the learning
process. If an unit does not receive inputs from
consecutive units, then the connections could be defined
separately with letter "r" or a period "." which specifies
an unmodifiable fixed weight value of zero. The same
macros has been used in the "biases" section which defines
the biases of the units in the network.
The "n-sample.spt" is the sinusoidal pattern file
which contains the input/output patterns of the network.
Each pattern in the file starts with a pattern name such as
"p.O" or pattern number 0, followed by one input sample,
ten current_state inputs from hidden units and finally
eleven desired output values.
The "n-sample.swt" contains the initial random
weights assigned to the connections in the network. The
weight values in this file are listed in order in which
they are defined in the n-sample.net file. Also all the
unit biases are included at the end of this file where
input and current_state unit biases are always set to zero.
The "n-sample.str" is the start file which sets the
parameters required in the back propagation learning
algorithm before program execution. The first two lines
tells the program to obtain network and input/output


33
informations from the n-sample.net and n-sample.spt files,
respectively. The next line set the number of iterations
that program executes each time the command "strain" (start
training) is activated. One epoch is defined as
representation of all input/output patterns for one
iteration. Next the error criteria threshold "ecrit" ,
the display and save levels are set to the desire values.
The mode "lgrain" or learning grain is set to pattern which
informs the program to update the weight values after each
pattern presentation. The learning rate "Irate" is set to
0.5 and parameter mji is set to 0.1 and finally the weight
values are obtained from n-sample.swt file.
The "n-sample.tem" is the template file which
describes the way various parameters or simulation output
results will be displayed on the CRT screen. In the define
section, the number of rows and columns to be used on the
screen (for example 47 rows and 131 columns) and the
location of each parameter with a "$" sign, is specified.
Then each parameter type and its field is declared in the
template specification part. For example the statement:
"ao vector 2 $ n activation h 3 100.0 42 21"
describes that "ao" is the output activation vector of
length 21 (unit 42 through 62) and its location is at the
nth occurrence of $ sign. This vector should be displayed
horizontally (h) with each activation field of three spaces


34
and a multiplication factor of 100, i.e. a value of 0.53
will be displayed as 53. This file is common to both
simulation runs of n-sample delay network.
In the first line the current epoch number,
total_Sum_square (tss) which results from adding the total
errors due to all patterns, gradient descent correlation
(gcor) that indicates the direction in which the searching
algorithm is moving, the sum square for each pattern (pss),
and the pattern name (cpname) is displayed. Other
important results such as hidden activations (ah), output
biases (bo), hidden biases (bh), output delta error (do) ,
hidden delta error (dh) and weight values are displayed in
the remaining lines.
The bp program updates the above parameters at the
end of each epoch and anytime tss value falls below the
desire error criteria or ecrit, it is assumed that the
problem has been solved and the network has learned the
necessary mapping to generate the desire outputs and
therefore the program will be terminated. Then after
issuing a "tall" command, the network output results for
each pattern can be viewed on the CRT by hitting the return
key.
In order to obtain the sinusoidal response, the bp
program can be activated by the following command:
"bp n-sample.tern n-sample.str"


35
The initial CRT displays for pattern 0 (p.O)
through pattern 20 (p20) are shown in appendix B. As can
be seen from these displays, the initial activations of
most output units (ao) are around o.5 which indicates the
net sum to each unit is close to zero. These activations
are not close to their desire or target activation values
(ta), so learning is turned on by the command "strain" and
as iterations are processed, the output activations of the
units will start to change toward their desire values.
After 247 iterations, the network has found the desire
mapping with an average error of about 5.2 percent per
sample. The final simulation results of p.O through p20
are included in appendix B. The desire outputs verses the
actual outputs of the n-sample delay network for sinusoidal
samples are shown in figure 13.
The impulse response network is then trained by
using the n-sample delay network desire outputs as inputs.
This network will learn the necessary mapping with an
average error of 0.18 percent per sample in about 150
iterations. The files required for this simulation and the
initial and final results for each pattern are included in
appendix B. The desire outputs verses actual outputs of
this network is shown in figure 14.
At this point, the impulse response network has
been trained and it contains the impulse response samples


36
internally. Therefore, with no learning the actual outputs
of the n-sample delay network are presented as inputs to
this network and the desire outputs verses actual outputs
are shown in figure 15. The files and final CRT displays
for patterns p.O through p20 are included in appendix B.
PNIT STEP RESPONSE
In previous simulation, the n-sample delay network
and impulse response network have been trained to generate
n-1 delay samples of a sinusoidal input and corresponding
filter outputs, respectively. Therefore this system should
be able to produce the filter output y(n) based on any given
input samples with.no training. Twenty-one samples of an
unit step signal are inputted to the n-sample delay network
serially. The output values will shift to the right by one
for every sample presentation and at the same time are fed
to'the impulse response network through switches SO-Sn-i.
The actual output verses desire output results of n-sample
delay and impulse response networks are shown in figures 16
and 17, respectively, and their files and displays are
included in appendix B. The impulse response network is
also presented with the desire outputs of n-sample delay
network as inputs and the results are shown in figure 18.
The files and displays of these patterns are also included
in appendix B.


Network Output
Figure 13: n-sample Delay Network Output for Sinusoidal Input Patterns
u>
-j


Network output
Figure 14: Sinusoidal Response of Impulse Response Network
with Desire Outputs of n-sample Delay Network as inputs.
LO
oo


Network output
Actual output
Desire Output
n
Figure 15: Sinusoidal Response of Impulse Response Network
with Actual outputs of n-sample Delay Network as inputs.
i>>
vo


Network output
1.11
1.0
0.9
0.8-
0.7-
0.6-
0.5-
0.4-
0.3-
0.2-
0.1*
T B
B B
a b
B B B
B
H Actual output
Desire Output
n
0.0 "I1it-1i'i1i1r
ii|i|i|iiiiii
0 2 4 6 8 10 12 14 16 10 20 22 24
Figure 16: n-9ample Delay Network Outputs for Unit Step Input Patterns.


Network output
Actual output
Desire Output
n
Figure 17: Unit Step Response of Impulse Response Network
with Actual outputs of n-sample Delay Network as inputs.


Network output
Figure 18: Unit Step Response of Impulse Response Network
with Desire outputs of n-sample Delay Network as inputs.
K>


43
SYSTEM IDENTIFICATION AND MODELING
In many engineering and control applications, a
system or a plant of unknown structure has observable input
and output signals. One way of obtaining knowledge about
the unknown system's dynamic response is to apply its input
to an adaptive neural network and to use its output as the
adaptive neural network's desired response. The general
architecture of this model is shown in figure 19.
plant unknown plant
input
*
neural network

Z
error signal
plant
output
Figure 19: System Identification Model
The FIR neural network develops an impulse response to
match that of the unknown plant since the neural network
and the plant develop similar outputs when driven by the
same input. In this manner, the dynamics of any plant can
be identified and modeled as above. Lets assume that the


44
unknown plant is a first order system as shown in figure
20.
Figure 20: The Plant Network
The transfer function of this plant can be found from
figure 21 as:
C (s) 1 R(S) 1 C (s)
R(S) TS+1 TS + 1
Figure 21: The Plant Transfer Function-
Assuming T=l, the unit step response of the plant is:
C(t) = 1 e-t/^ = 1 et t>0
Therefore, in order to determine the dynamic behaviors of
this unknown plant, a system identification architecture as
in figure 19 is modeled. A unit step is provided as input
to both the plant and neural network. Plant modeling is
accomplished by the criteria of minimizing the error


45
between the plant and neural network outputs. The output
responses of the plant and the FIR neural network are shown
in figure 22 and the simulation files are included in
appendix B.
Figure 22: Output Responses of the Plant and Neural Network


REFERENCES
[1] J. T. Tou and R. C. Gonzales, Pattern Recognition
Principles (Addison Wesley Publishing Company, 1974).
[2] F. Rosenblatt, Principles of Neurodynamics (Spatan
Book, NJ, 1962),
[3] G. Widrow and M. E. Hoff, Adaptive Switching Circuits
(Institute of Radio Engineers, Western Electronics
Show and Convention, Part 4, 1960), P. 96-104.
[4] S. Grossberg, Adaptive Pattern Classification and
Universal Recording. Part I, Parallel Development and
coding of Neural Feature Detectors. (Biological
Cybernetics, 23, 1976), P. 121-134.
[5] D. E. Rumelhart and J. L. McClelland, An Interactive
Activation Model of Context Effects In Letter
Perception,. Part 2, The Contextual Enhancement Effect
and some Tests and Extension of the Model.
(Psychological Review, 89, 1982), P. 60-94.
[6] G. E. Hinton, T. J. Sejnowski and D. H. Ackley,
Boltzman Machines; Constraints Satisfaction Networks
that Learn. (Carnegie-Mellon University, Department
of Computer Science, Pittsburgh, PA, 1984).
[7] D. E. Rumelhart and J. L. McClelland, Parallel
Distributed Processing. Volume 1, Foundations,
(The MIT Press, 1987).
[8] G. Widrow and S. D. Stearns, Adaptive Signal
Processing. (Prentice-Hall, NJ, 1985).
[9] A. V. Openheim and R.W. Schafer, Digital Signal
Processing (Prentice-Hall, NJ, 1975).
[10] M. I. Jordan, Attractor Dynamics and Parallelism
in a Connectionist Sequential Machine, Proceedings
of the Eight Annual Conference of the Cognitive
Science Society (Hillsdjale, NJ: Lawrence Erlbaum
Associates, 1986). !


47
[11] D. E. Rumelhart and J. L. McClelland, Exploration
in Parallel Distributed Processing. (The MIT Press,
1988) .


APPENDIX A
COMMAND SET
In the bp program an interactive command set exist
which can be used to start the simulation or to activate
various functions during the simulation. The commands used
in the simulations of this report are listed below.
For more informations refer to [11].
bp
Activates the bp program. This
command must follow by a .tern and
a .str file, respectively.
strain
Starts the training for all the
patterns represented to the
network. This command executes
sequentially for nepoch
iterations.
tall
Tests all patterns on the pattern
file one by one, and pauses after
each pattern presentation.
Learning is turned off when tall
is active.
quit
clear
Terminates the program.
Clears the screen.
Directs the program to extract
network informations from the
filename file.
get network filename


49
get pattern filename Directs the program to extract pattern informations from the filename file.
get weight filename Directs the program to extract weight and bias informations from the filename file.
save weight filename Directs the program to save the weights and biases of the network in the filename file.
set ecrit value Sets the error criteria value to value which determines when to stop the program.
set lflag value This command sets the variable lflag to value which indicates when learning is on. A 1 will activate the learning and 0 will deactivate.
set nepoch value This command will set the number of epochs to value that will be executed after a strain command.


APPENDIX B
SIMULATION FILES AND DISPLAYS
In this section all the simulation files and the initial
and final output results of n-sample delay network and
impulse response network are shown. In the case of
training a network, all initial states of the network
before learning and all final results after the training
for each of the twenty-one sinusoidal patterns are shown.
Finally the response of the unit step input patterns for
both networks are displayed.


nsample.str file for sinusoidal pattern simmulation
get network nsample.net
get patterns nsample.spt
set nepochs 100
set ecrit .1
set dlevel 3
set slevel 1
set lflag 1
set mode follow 1
set mode lgrain pattern
set param Irate .5
set param mu 0.1
get weight nsample.wts


nsample.net file
definitions
nunits G3
ninputs 21
noutputs end network: i 21
%r 21 1 0 1
%r 22 1 0 2
%r 23 1 1 2
%r 24 1 2 2
%r 25 1 3 2
%r 26 1 4 2
%r 27 1 5 2
%T 26 1 6 2
%T 29 1 7 2
%r 30 1 e 2
%T 21 1 9 2
Mr 32 1 10 2
%T 33 1 11 2
%T 34 1 12 2
%T 35 1 13 2
%r 36 1 14 2
%r 37 1 15 2
& /<>* 38 1 16 2
%r 39 1 17 2
/a r 40 1 ie 2
%T 41 i 19 2
% 42 1 0 22
r. / /t 4 5 i i 42
O' /C - 44 i 2 42
o' n 4 5 i 3 "l J
/ /V 46 i 4 42
r*
/ /e 47 i 5 42
O' /* 48 i 6 42
.
O' JV 49 i 1 42

/V 50 i 8 4 * 1 L
*' 51 i 9 42
B/ /. 52 i 10 4 2

O' ft 12 2 x to
. .
4'

54 1 1
o


r
X 56 1 14 42
r.......................r........................
% 57 1 15 42
...............................................r
% 58 1 16 42
r.......................r......................r
% 59 1 17 42
........................r......................r
% GO 1 18 42
........................r.................... r
% 61 1 19 42
...............................................r
% 62 1 20 42
........................r......................r
end
biases:
%r 21 42
end


nsample.tem file
define: layout 47 131
epoch $ tss S pss s gcor S cpname ) S in S
ta S ah S ao S do $ bill S bh2S bo IS bo2S
v21 S w32S w042S w42S wO 5 3 S w5 3s w52 5 s
w22s w33S w04 3 S w4 3S w4 24S w054 S w54 S w5 35 s
w2 3 5 w3 4S U-044S w44s w4 34S w055S w5 5s w54 5 s
w24 S w35S w04 5S w4 5S w444$ w05G$ w5G S w555 s.
w25S w3GS w04 6S w4 6S w4 54S w057s wo r S w565 s
w26$ w37S w04 7S w4 7$ w4 G4 $ w058S w58S w57 5 s
w27s w38S w04 8S w48S w4 7 4 S w059S w5 9S w585 s
w28S w39S w049s w49S w484S wOGOS w60S w596 s
u29S w40S w050s w50S w495$ w06 1 S w6 1 S wG0 6 s
w30s w4 IS w051S w5 1 S w505S w062S wG2S wGIG s
w31S w052S w52S 515S
end epochno variable 1 S n epochno 6 1.0
tss floatvar 1 S n tss 6 1.0
gcor floatvar 2 S 3 gcor 6 1.0
cpname variable 2 s n cpname -4 1.0
pss floatvar o U s 2 pss 6 1.0
in vector 2 s 5 activation h 3 100.0 0 1
ah vector 2 s 7 activation h 3 100.0 21 21
ao vector 2 s n activation h 3 100.0 42 21
ta vector 2 s 6 target h 3 100.0 C 2 1
do vector 2 s 9 delta h 3 100.0 42 21
bhl vector 2 s n bias h 4 100.0 21 11
bh2 vector 2 s n bias h 4 100.0 32 10
bo 1 vector 2 s n bias h 4 100.0 4 2 11
bo2 vector 2 s n bias h 4 100.0 53 10
w21 matrix 3 e n weight h 4 100.0 21 1 0 1
w32 matrix 3 s n weight h 4 100.0 32 1 10 2
w04 2 matrix 3 s n weight h 4 100.0 42 1 0 1
w4 2 matrix 3 s n weight h 4 100.0 42 1 21 1
w05 3 matrix 3 s n weight h 4 100.0 53 1 11 1
5 3 matrix 3 s n weight h 4 100.0 53 1 32 1
v525 matrix 3 s n weight h 4 100.0 5 3 1 52 1
w22 matrix 3 s n weight h 4 100.0 22 1 0 2
w33 matrix o s n weight h 4 100.0 o o O U 1 11 ~2
i-04 3 matrix o J s n weight h 4 100.0 43 1 1 1
v4 3 matrix 2 s n weight h 4 100.0 43 1 22 n
>-4 24 matrix 3 s n weight h 4 100.0 43 1 4 2 1
-054 matrix 2 s n weight h 4 100.0 54 i 12 1
-54 matrix 3 s n weight h 4 100.0 54 1 3 *' 1
3 u matrix 3 s n weight h 4 100.0 54 1 s o 1
<-2 3 matrix 3 s n weight h 4 100.0 23 1 1


w044 matrix 3 $ n weight h 4 100.0 44 i 2 1
w44 matrix 3 $ n weight h 4 100.0 44 i 23 1
w434 matrix 3 s n weight h 4 100.0 44 i 43 1
w05 5 matrix 3 s n weight h 4 100.0 55 i 13 1
w5 5 matrix 3 s n weight h 4 100.0 5 5 i 34 1
w54 5 matrix 3 s n weight h 4 100.0 5 5 i 54 1
w24 matrix 3 s n weight h 4 100.0 24 i o o 1- 4a
w3 5 matrix 3 s n weight h 4 100.0 35 i 13 2
w04 5 matrix 3 s n weight h 4 100.0 4 5 i 3 1
w45 matrix 3 s n weight h 4 100.0 45 i 24 1
w4 4 4 matrix 3 s n weight h 4 100.0 4 5 j 44 1
w056 matrix 3 J n weight h 4 100.0 56 i 14 1
w5G matrix 3 c n weight h 4 100.0 5G i 35 1
u-55o matrix 3 s n weight h 4 100.0 56 i 5 5 1
w25 matrix 3 £ n weight h 4 100.0 2 5 i 3 2
w36 matrix 3 s n weight h 4 100.0 36 i 14 2
w04 6 matrix 3 g n weight h 4 100.0 4 G i 4 1
w4 6 matrix 3 s n weight h 4 100.0 4 G i 25 1
w4 54 matrix 3 c n weight h 4' 100.0 46 i 45 1
w05 7 matrix 3 s n weight h * 100.0 57 i 15 1
k5 7 matrix 3 s n weight h 4. 100.0 5 7 i 36 1
k565 matrix 3 s n weight h 4 100.0 57 2 56 1
w26 matrix 3 s . n weight h 4 100.0 26 1 4 2
w37 matrix 3 s n weight h 4 10C.0 37 1 15 2
w04 7 matrix 3 s n weight h 4 100.0 47 1 5 1
w4 7 matrix 3 s n weight h 4 100.0 4 7 l 26 1
w4G4 matrix 3 s n weight h 4 100.0 47 1 46 1
w058 matrix 3 s n weight h. 4 100.0 58 1 16 1
w58 matrix 3 s' n weight h 4 100.0 58 1 37 1
w57 5 matrix 3 s n weight h 4 100.0 58 1 5 7 1
w27 matrix 3 s n weight h 100.0 27 1 5 2
w38 matrix 3 s n weight h 4 100.0 38 1 1G 2
w04 8 matrix 3 s n weight h 4' 100.0 48 1 5 1
w4 8 matrix 3 s n weight h 4 100.0 48 1 27 1
w4 74 matrix 3 s n weight h 4 100.0 48 1 47 1
w059 matrix O s n weight h t 100.0 59 1 17 1
w59 matrix 3 s n weight h 4 100.0 59 1 38 1
w585 matrix 3 s n weight h 4 100.0 59 1 58 1
v28 matrix 3 s n weight h 4 100.0 28 1 6 2
w39 matrix O o s n weight h 4 100.0 39 1 17 2
v04 9 matrix 2 s n weight h 4' 100.0 49 1 7 1
w4 9 matrix 3 s n weight h 4 100.0 49 1 28 1
w4 84 matrix 3 s n weight h 4 100.0 49 1 48 1
w060 matrix 2 c n weight h 4 100.0 60 1 18 1
-60 matrix 3 s n weight h * 100.0 60 1 39 1
v596 matrix 3 s n weight h 4 100.0 60 1 50 1
w29 matrix 3 c n weight h 100.0 29 1 7 2
w4 0 matrix 3 e n weight h 100.0 40 X -* 00 *0
-050 matrix 3 e n weight h 4 100.0 50 1 8 1
voO matrix 3 s n weight h 1 n 100.0 50 2 29 1
w4 9 5 matrix 3 c- r; we ight h 4 100.0 50 1 49 1
vOGl matrix 0 s ri . weight h 4 100.0 61 2 19 1
u-61 matrix 3 g n weight h 4 100.0 61 1 40 1
-SO 6 matrix 2 s n weight' h 4 100.0 61 1 60 1
-30 matrix o sj e n weight h 4 100.0 30 1 £
v4 1 matrix 3 s n weight h 4 100.0 -i - i 19 2
-051 matrix 3 s n : igh t h n 100.0 r i j i 1 9 1


w51 matrix 3 n weight h 4 100.0 51 30 1
w505 matrix 3 s n weight h 4 100.0 51 1 50 1
w062 matrix 3 s n weight h 4 100.0 G2 1 20 1
wC2 matrix 3 c n weight h 4 100.0 62 1 41 1
w616 matrix 3 s !l weight h 4 100.0 62 1 61 1
w31 matrix 3 s n weight h 4 100.0 31 1 9 2
w0 52 matrix 3 s n weight h 4 100.0 52 1 10 1
w52 matrix 3 G n weight h 4 100.0 52 1 31 1
w515 matrix 3 s n weight h 4 100.0 52 1 51 1


nsample.spt file
p.O 0.500 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.500 0.500 0.206 0.024 0.024 0.206 0.500 0.793
0.975 0.975 0.793 0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975
0.975 0.793
p.l 0.793 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.793 0.500 0.500 0.206 0.024 0.024 0.206 0.500
0.793 0.975 0.975 0.793 0.500 0.206 0.024 0.024 0.206 0.500 0.793
0.975 0.975
p.2 0.975 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.975 0.7930.500 0.500 0.206 0.024 0.024 0.206
0.500 0.793 0.975 0.975 0.793 0.500 0.206 0.024 0.024 0.206 0.500
0.793 0.975
p.3 0.975 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.975 0.975 0.793 0.500 0.500 0.206 0.024 0.024
0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.206 0.024 0.024 0.206
0.500 0.793
p.4 0.793 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.793 0.975 0.975 0.793 0.500 0.500 0.206 0.024
0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.206.0.024 0.024
0.206 0.500
p.5 0.500 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.500 0.793 0.975 0.975 0.793 0.500 0.500 0.206
0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.206 0.024
0.024 0.206
p. 6 0.206 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.500
0.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.206
0.024 0.024
p.7 0.024 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500
0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500
0.206 0.024
p. 8 0.024 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793
0.500 0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793
0.500 0.206
p.9 0.206 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975
0.793 0.500 0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975
0.793 0.500
pi0 0.500 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975
0.975 0.793 0.500 0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975
0.975 0.793
pll 0.793 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -53 -59 -60 -61 0.793 0.500 0.206 0.024 0.024 0.206 0.500 0.793
0.975 0.975 0.793 0.500 0.500 0.206 0.024 0.024 0.206 0.500 0.793
0.975 0.975
p!2 0.975 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.975 0.793 0.500 0.206 0.024 0.024 0.206 0.500
0.793 0.975 0.975 0.793 0.500 0.500 0.206 0.024 0.024 0.206 0.500
0.793 0.975
pi3 0.975 -42 -4 3 -44 -4 5 -46 -4 7 -48 -49 -50 -5 1 -52 -53 -54 -55 -
-57 -58 -59 -60 -61 0.975 0.975 0.793 0.500 0.206 C.C24 0.024 0.206
56
56
I
56
56
56
56
5 6
56
56
56
56
56
56
56


0.500 0.793 0.975 0.975 0.793 0.500 0.500 0.206 0.024 0.024 0.206
0.500 0.793
pi4 0.793 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -:52 -53 -54 -55 -56
-57 -58 -59 -60 -61 0.793 0.975 0.975 0.793 0.500 0.206 0.024 0.024
0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.500 0.206 0.024 0.024
0.206 0.500
pi5 0.500 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -56
-57 -58 -59 -60 -61 0.500 0.793 0.975 0.975 0.793 0.500 0.206 0.024
0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.500 0.206 0.024
0.024 0.206
pl6 0.206 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -56
-57 -58 -59 -60 -61 0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.206
0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.500 0-206
0.024 0.024
pi7 0.024 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -56
-57 -58 -59 -60 -61 0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500
0.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793 0.500 0.500
0.206 0.024
pi 8 0.024 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -56
-57 -58 -59 -60 -61 0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793
0.500 0.206 0.024 0.024 0.206-0.500 0.793 0.975 0.975 0.793 0.500
0.500 0.206
pi9 0.206 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -56
-57 -58 -59 -60 -61 C.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975
0.793 0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975 0.793
0.500 0.500
p20 0.500 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -56
-57 -58 -59 -60 -61 0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975
0.975 0.793 0.500 0.206 0.024 0.024 0.206 0.500 0.793 0.975 0.975
0.793 0.500


n-samplt aeiay network dis
erns
rs:i:a
p.ays for sinusoidal input patterns
p to push/b to break/ to continue:
cisp/ exam j ' get/ save/ set/ cleer cycle do log newstart ptrain qu
reset run strain tall test
epoch 0 ts s 2.6242 pss 2.6242 gcor 0.0000 cpname p.O i n 5 0
ta 50 50 2 0 2 2 20 50 79 97 97 79 50 20 2 2 20 50 79 97 97 79
ar. 4 4 4 C 5 0 42 44 51 49 Cl 4 G 5 1 42 4 0 47 4 6 40 51 5 G CO 43 56 1 i
ao 45 52 5 2 G5 47 52 57 45 50 G1 41 58 0E 3G 37 DC- oo 42 49 53 39
do 1 0 - £-14 10 -6 -L 9 12 9 6 -2 -9 -6 -7 -9 -2 7 12 9 9
bhl -36 -1 3 -29 -22 G 0 47 -15 4 -25
bh2 -3G -9 -6 -15 C 27 44 -2C 35 -22
bo 1 -22 2C 0 19 -:2 23 45 -G 9 21 -3G
bo2 15 40 - 7 -44 42 4 -26 1 22 -30
w21 3G w' 2 -22 6 1.-042 -C w42 19 w0f;2 23 w o 3 19 w 2 25 2 9
w22 -27 4G v 2 3 -1 15 w0 4 3 12 w4 2 -44 w424 3 w0 54 1 3 w 5 4 4 9 wo 3 5 2 2
w23 1A -33 v34 4 0 4 2 w 0 4 4 4 i v 4 4 0 v 4 3 4 1C w055 -2C w55 - 1 2 WOl 1 -46
w 2 4 1 £. 3 3 w3 5 3 9 i o 29 v t :. 4 3 w 4 4 4 4 i w0 5G - 3 6 \ o G - 2 c> w 0 5 0 O O
w2 5 -16 10 w3G .43 -21 w04G 26 w4G -25 w454 22 v 0 5 7 - 13 w5 7 17 woC o _ -
w2C -C 9 v3 7 3 -9 w04 7 0 w47 14 w4G4 -42 w056 6 w5 8 5 wo i 5 2 5
v2 7 -5 4 w38 22 -14 w04E 18 w.;f: -34 w4 74 5 w059 -24 wf 9 i i wo 6 o - 3C
w o g r> r. v n i:3 9 -20 32 w04 9 -21 w49 17 w484 -3G wOCO 13 wCO 34 5 9 G -4 5
w29 39 -20 w4 0 39 -31 w050 -41 w50 -44 w495 30 wOCl 18 wCl - 20 wGOG 11
w30 17 -4 G w4 1 37 -19 w051 34 wal 0 w505 30 w0G2 2 wG2 24 wG 1 G -43
w31 -40 -40 w052 22 w52 -49 w5I5 42
p tc push/b to b reak/ to continue:
c i s p / e a ni / g e t/ save/ set/ clear cycle do log news tart strain qu
resc-t ru;. stra in tail test
epoc h 0 tss 5 . 5 5 6 3 pss 2.7421 SCOT 0. 0000 cpn time p. i i r 79
ta <9 j0 a0 20 2 2 20 50 79 97 97 79 50 20 2 2 20 50 79 97 97
ah 47 49 48 4 0 4 3 51 4 9 66 49 47 33 36 49 62 r 2 5 3 5 5 C 3 44 59 46
ao 45 53 47 GC 52 52 GO 4 2 4 5 G 5 4 7 G1 70 32 3 4 59 5G 29 50 5G 39
do 3 C 0- 10-11 - 12 -9 2 9 6 12 4 3 2 -6- 14 .-9 i 7 8 13
bh: -38 -1 3 -29 _ o o C 0 47 - 15 4 - 2 5
bh 2 -36 -9 - c : 5 6 27 44 -26 o r, o 9
bo 1 -22 2G f, } 9 -12 23 45 -6 9 31 - 3G
bo 2 15 4 0 - 17 -4 4 n w 4 -28 1 22 -30
w 21 3C w 32 2 2 e -04 2 -G v4 2 1 9 w05 3 2 2 1.1 9 s*5 2 r oc
v 2 3 -27 46 u O' _ 1 5 w04 3 12 w4 3 -44 w424 3 wC54 *3 V 4 9 w5 3 5 rt O
w23 14 -33 w *5 4 r Cl ~l . / o n _ w044 -47 w44 0 w s j - 1G w 0 a a -20 w 5 5 - 2 w 5 4 5 Sr.
w24 IS -33 w 35 29 / w04 5 29 w4 5 .4 3 w4 4 4 4 / w0 5 6 -3£ v56 - S V Ll w D *> o
w25 -16 10 v 36 4 2 - 21 w04G 26 w4 6 - 2 5 w4 54 O *. w 9 a 7 _ i r. w5 7 7 >. 5 6 5 _ 2 "!
---26 -6 9 w 2 7 2 _ Q w047 0 w47 14 w464 -42 w0 5S c w 5 e 5 'Ll 7 0 2 5
..A* ( - 5 4 - 3 r. Z 2 -1 4 .-0 4 6 IS w4S -54 w474 £ w059 - 2 4 v 5 S 3 2 - 5* S 5 - 2 C
.. c > b u n n ? 9 j 2 9 2 0 2 2 1.-04S -2; w-4 9 17 w4 64 -35 wOGO 2 v wCO 3-i -596 -4 5
. r r 39 -20 i. 4 < C: IV w -: -050 -4 1 i.-50 -44 - 4 95 1 r wOCl 2 £ ioC : -20 .COG . i
.. *. r. w U 17 4 5 41 3 7 -: f; w 0 51 34 w 51 0 a 0 o 30 wOG2 2 wG2 2-i .2 : G - 4 2
V 2 I -40 -40 r.rrt r* o ,. r, *" 4 6 i.1: 1 : 4 2
i


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 0 tss 8 . 1736 pss 2. 6073 gcor 0. 0000 cpname P-2 in 9 7
ta 97 79 50 50 20 2 2 20 50 7 9 97 97 79 50 20 2 2 20 50 79 97
ah 49 48 47 41 44 51 49 69 50 4 7 32 38 49 63 52 5 3 5 5 63 4 3 58 46
ao 45 53 46 68 52 52 60 42 45 65 47 62 70 32 34 59 5 7 39 51 56 39
do 13 6 0-4 -6- 12-13 -5 1 4 13 8 1 3 -2- 14-12 -4 0 4 13
bhl -38 -1 3 -29 -22 6 0 47 -15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 -22
bol -22 26 5 19 -13 23 45 -G 9 31 - 36
bo2 15 40 -17 -44 43 4 -28 1 22 -30
i,-21 36 k32 -22 8 i,-04 2 -6 i,-4 2 19 u-05 3 23 w53 19 w525 29
w22 -27 46 w33 -1 1 5 w04 3 12 w43 -44 i.-4 24 3 w054 13 t,-54 49 v53 5 > o
w23 14 -33 i.-34 4G 42 u-04 4 - 4 7 w4 4 0 i,-4 34 16 i,-0 5 5 -20 i,-5 5 - 1 2 i,-5 4 5 -48
w24 18 -33 i,-35 39 - 7 w04 5 29 w4 5 4 3 i,-44 4 47 i,-056 -38 waG - 28 w o 0 5 22
w25 -18 10 t,-36 4 3 -21 w04 6 28 i,-4G -25 w4 54 22 i,-057 -13 w5 7 17 w56 5 -21
w26 -6 9 w37 3 -9 w04 7 0 w4 7 14 w464 -42 i,-058 8 !,* y 8 5 l." 5 7 5 25
i,-27 -5 4 i,-38 32 -14 i,-04 e 18 w48 -34 w4 7 4 5 w059 -24 w5 9 33 w585 -36
w28 22 32 i,-39 -20 32 w04 9 -21 w4 9 17 V.-4 84 -36 1.-060 13 wGO 34 w596 -45
i>29 39 -20 w4 0 39 -31 i,-0 50 -4 1 u-50 -44 1-4 95 30 i,-061 18 wGl - 20 wCOC 11
w30 17 -46 u-41 37 -19 w05 1 34 w51 0 w505 30 w062 2 w62 24 i.-GIG -43
u-31 -40 -40 w052 22 i,-52 -4 9 w5 15 42
p to push/b to break/ to continue:
aisp/ exam/ get/ save/ set/ clear cycle do log newstart p (train qui
rese t run strain tall test
epoch 0 tss 10.821 pss 2 . 6478 gcor 0. 0000 cpname p.3 r. 9 7
ta 97 97 79 50 50 20 2 2 20 50 79 97 97 79 50 20 2 2 20 50 79
ah 49 4 8 4 7' 41 44 51 49 69 50 4 7 32 38 50 64 53 53 55 63 4 3 58 46
ao 4 5 53 46 67 52 52 60 42 45 65 47 62 70 31 34 59 57 39 51 56 39
do 13 11 7-3 0 -8-12 -10 -6 - 2 8 8 4 10 4-10-12 - 7 -7 -2 9
bhl -38 -1 3 -29 -22 6 0 47 - 15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 -22
bo 1 -22 26 5 19 -13 23 45 -6 9 31 - 36
bo2 15 40 -17 -44 43 4 -28 1 22 -30
i.-21 36 i,-32 -Oil 8 w04 2 -6 w4 2 19 w053 23 l,* y 3 19 w525 29
w22 -27 46 w33 -1 15 1.-04 3 12 w4 3 -44 w4 24 3 w054 13 i.-54 49 w 5 3 5 22
i.-2 3 14 -33 i-34 46 42 i-044 -4 7 w44 0 1,4 34 16 w055 -20 v5 5 -12 w54 5 -48
v24 18 -33 w35 39 - 7 v04 5 29 w4 5 43 1.-444 47 w056 -38 w56 -28 w5 5 5 22
w2 5 -18 10 w36 43 -21 1.-04 6 28 i.-4 6 -25 1.-4 54 2 2 wO 5 7 -13 w57 17 v565 -21
v-26 -6 9 i." 3 7 3 -9 i-04 7 0 w4 7 14 v464 -42 w058 8 w58 5 w 5 7 5 25
w27 - 5 4 i-38 32 -14 i,-04 8 18 i.-4 8 -34 i,-4 7 4 5 w0 5 9 -24 i,-59 33 1,-585 -36
i.-2 8 22 32 i.-3 9 -20 32 w04 9 -21 1-4 9 17 v4 34 -36 I.-060 13 w6 0 34 w5 96 -45
w29 39 -20 w40 39 -31 w050 -41 w50 -44 1.-495 30 1,-OGl 18 i-61 -20 w606 11
i,-30 '17 -4 6 w.'l 27 _19 ..-051 34 w51 0 '.-505 30 1.-062 t,-C2 O 1 ^ -I i:61C- -43
1.-31 -40 -40 u-052 22 i.-52 -49 w515 1 o *1 *-


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 0 tss 13.525 pss 2. 7040 gcor 0. 0000 cpname P-4 in 79
ta 79 97 97 79 50 50 20 2 2 2 0 50 79 97 97 79 50 20 2 2 20 50
ah 4 7 50 47 41 44 51 49 69 50 4 7 32 38 50 64 5 3 53 55 63 43 58 46
bo 4 5 53 46 67 52 52 60 42 45 65 47 62 70 31 34 59 57 39 51 56 39
do 8 11 12 2 0 0 -8-10- 11-10 0 4 4 14 10 -2 -8 -7- 12 -9 o
bhl -38 -1 3 -29 -22 6 0 47 - 15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 -22
bol -22 26 5 19 -13 23 4 5 -6 9 31 - 36
bo2 16 40 -17 -44 43 4 -28 1 O'} -30
w21 36 w32 -22 8 w04 2 -6 w4 2 19 w0t3 23 w5 3 19 w525 29
k22 -27 46 v3 3 -1 15 w04 3 12 w4 3 -44 w4 24 3 w054 13 1.-54 49 w o 3 o 22
w23 14 -33 v3 4 4 6 4 2 w04 4 -47 w4 4 0 w4 34 16 w05 5 -20 w 5 5 - 12 w54 5 -48
w24 IS -33 w3 5 39 -7 v045 29 v4 5 43 w4 4 4 47 w05G -38 w 5 G - 28 w 5 5 5 0 o
w25 -18 10 w36 43 -21 w046 28 w4 6 -25 w4 54 o o w05 7 -13 w5 7 17 w5G5 -21
w26 -6 9 w37 3 -9 w047 0 w4 7 14 w4 64 -42 w058 8 w58 5 w 5 7 5 25
w27 -5 4 w38 32 -14 w048 18 w4 8 -34 w4 7 4 5 w059 -24 w59 33 w585 -3G
w28 22 32 w39 -20 32 w049 -21 w4 9 17 w4 84 -36 w060 13 w60 34 w596 -45
w29 39 -20 w4 0 39 -31 w050 -41 w50 -44 w4 95 30 w061 18 wGl - 20 w606 11
w30 17 -46 w41 37 -19 w051 34 w51 0 w505 30 w062 2 w62 24 w616 -43
w31 -40 -40 wO o 2 22 w52 -49 wa 15 42
p to push/b to breah/ to con tinue:
cisp/ exam/ get/ save/ set/ clear cycle do log newstart I >train qu:
reset run strain tall test
epoi ch 0 tss 16.365 . pss 2. 8401 gccr 0. 0000 cpname p. 5 in 50
ta 50 79 9 7 97 79 50 50 20 o 2 20 50 79 97 97 79 50 20 2 2 20
ah 44 52 4 7 41 44 51 4 9 69 50 4 7 32 38 50 64 5 3 53 55 63 43 58 46
ao 45 52 46 67 52 52 60 42 45 65 47 62 70 31 34 59 57 39 51 56 39
do 1 7 1 3 6 6 0 -1 -6- 11-14 -6 -2 0 14 14 4 -1 -3- 12-12 -4
bhl -38 -1 3 -29 -22 6 0 A -15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 -22
bol -22 26 5 19 -13 23 45 -6 9 31 - 36
bo2 15 40 -17 -44 43 4 -28 1 22 -30
w21 3 6 w32 -22 8 w04 2 -G w4 2 19 1--053 23 w5 3 19 w 5 2 5 29
w22 -27 46 w33 -1 1 5 u-04 3 . 12 w4 3 A A w4 24 3 w054 13 w54 49 w53 5 22
..-23 14 -33 i.-3 4 4 6 42 w0 4 4 ^ 1 w4 4 0 w4 34 16 w055 -20 w55 -12 w54 5 -48
w24 18 -33 w35 39 -7 w04 5 29 w4 5 43 w4 4 4 47 w056 -38 w56 -28 w5 5 5
w25 -18 10 w36 4 3 -21 1,-04 6 28 w4G -25 i,-4 5 4 2 2 wO 57 -13 w57 17 i,-565 -21
w26 -6 9 w37 3 -9 w04 7 0 w4 7 14 w464 -42 w05B 8 w58 5 w 5 7 5 25
w27 -5 4 w38 3 2 -14 1.-04 8 18 w48 -34 1.-4 7 4 Cl 1.-059 -24 w59. 33 w585 -36
w28 22 32 w39 -20 32 I.-04 9 -21 w4 9 17 w4B4 -36 wOGO 13 w60 34 w596 4 5
w29 39 -20' w40 39 -31 1.-050 -41 i.-50 -4-1 w4 95 30 w061 18 i.-61 -20 wG06 11
w30 17 -46 i.-41 37 _ 1 o 1.-051 O 1 *1 i.-5 1 0 w505 30 1-062 2 w62 24 u-61 C -4 3
w31 -40 -40 1,05 2 r> a .1 C i.-52 -49 i,- zj 15 .1 *l .


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cj-cle do log newstart ptrain quit
reset run strain tall test
epoch 0 tss 19.291 pss 2.9260 gcor 0.0000 cpname p.6 in 20
ta 20 50 79 97 97 79 5C i 50 20 2 2 20 50 79 97 97 79 50 20 2 o
ah 42 54 47 41 44 51 49 i 69 50 4 7 32 38- 50 64 53 53 55 63 43 58 45
ao 46 52 45 57 52 52 6C i 4 2 4-5 6 5 47 62 70 31 34 59 57 39 51 56 39
do -6 0 8 6 10 6 -2 1 -7-15-12-10 -5 10 13 9 5 3 -7-12 -8
bhl -38 -1 3 -29 -22 6 0 47 -15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 _n n
bo 1 -22 26 5 19 -13 23 45 -5 9 31 - 36
bo2 15 40 -17 -44 43 4 -28 1 2 -30
w21 36 w32 _ 8 w04 2 -5 w4 2 19 w053 2 3 w53 19 w525 29
w22 -27 46 w33 -1 15 w043 12 k43 -44 w4 24 3 w054 13 w54 49 w53 5 r. r.
v23 14 -33 w34 46 42 w04 4 -4 7 w4 4 0 w4 34 16 w05 5 -20 w5 5 - 12 w54 5 -48
w24 18 -33 w35 39 - 7 u-04 5 29 w4 5 4 3 w4 4 4 4 7 w056 -38 w 5 6 - 28 w55 5 22
w25 -18 10 w3 6 43 -21 w04 6 28 w46 -25 w4 54 22 w05 7 -13 U 0 1 17 w 5 6 5 -21
w26 -6 9 w37 3 -9 w04 7 0 w4 7 14 w4 54 -42 w0 58 e w58 5 w57 5 25
w£7 -5 4 w38 3 0 -14 w048 ie w48 -34 w4 7 4 5 w059 -24 w5 9 33 w 5 8 5 -36
w28 22 32 iv39 -20 32 w04 9 -21 w4 9 17 w4 84 -36 wO60 13 wCO 34 w596 -45
w29 39 -20 w40 39 -31 w050 -41 w50 -44 w4 95 30 w061 18 w61 - 20 w6 06 1 1
w30 17 -46 w4 1 37 -19 w051 34 w51 0 w505 30 k062 2 u62 24 w616 -43
w31 -40 -40 w052 22 w52 -49 w515 42
p tc push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 0 ss 22.093 pss 2 . 8023 gcor 0. 0000 cpname p. 7 i: . o
ta 2 20 50 79 97 97 79 50' 50 2 0 2 2 20 50 79 97 97 79 50 20 o
ah 40 55 47 41 44 51 49 69 50 4 7 32 38 50 64 53 53 55 63 43 58 4 5
ao 46 52 46 67 52 52 60 42 45 65 47 62 70 31 34 59 57 39 51 56 39
do -10 -7 1 o U 9 11 4 i 0-1 1-12- 14- 10 4 9 9 9 0 -7 -8
bhl -38 -1 3 -29 -22 6 0 47 - 15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 -22
bo 1 -22 26 0 19 -13 23 45 -6 9 31 - 36
bo2 15 40 -17 -44 43 4 -28 1 22 -30
w21 36 w32 -22 8 w04 2 -6 w4 2 19 w053 23 v53 19 wo 25 29
w22 -27 46 w33 -1 15 w04 3 1 w4 3 -44 w4 24 3 w054 13 w54 49 w5 3 5 o o
w23 14 -33 w u 4 46 4 2 w 0 4 4 -47 w44 0 w4 3 4 16 w055 -20 w5 5 - 12 w54 5 -4 S
w2 4 18 3 3 w3 5 39 -7 w0 4 5 29 w4 5 43 w44 4 47 w056 -38 w 5 6 - 28 wo 55 22
w25 -1 8 10 w36 43 -21 w04 6 23 w4 6 -25 w4 5 4 22 w05 7 -13 v5 7 17 w5C5 - 21
w26 -6 9 w3 7 3 -9 w04 7 0 w4 7 14 w4 64 _ J o w058 8 w58 5 w5 7 5 0 s
w27 "0 *i w38 32 -1 4 u-048 18 w4 8 -34 w4 7 4 5 w05 9 -24 wo 9 33 w5S5 -36
w28 9 0 'J 9 w39 -20 32 w04 9 -21 w4 9 1 1 w4 84 -36 w060 13 w60 34 wo 9 6 -45
w 2 9 29 -20 w4 0 39 -21 w050 -4 1 w50 -44 w4 95 30 wOGl 13 w61 - 20 w606 1!
w30 17 -46 w4 1 2 7 1 9 ;.*05 1 34 wo 1 r> w50 5 30 w062 o wC2 24 w516 - 4 3
-..-31 1 Jn 0 1 o w052 o 2 wo 2 wo 1 5 A O


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain qu
reset run strain tall test
epoch 0 tss 2 4.776 pss 2. 6823 gcor 0. 0000 cpname p.8 in 2
ta 2 2 20 50 79 97 9 7 79 50 50 20 O 2 20 50 7 9 9 7 97 7 9 50 20
ah 4055 48 41 44 51 49 69 50 47 32 38 50 64 53 5 3 5 5 63 4 3 58 46
ao 46 52 46 67 52 52 60 42 45 6 5 47 62 70 31 34 59 5 7 39 51 5 C 39
do -11-12 -6 -3 5 1 1 8 9 0 - 4 -7- 14- 13 -2 3 5 8 13 7 -1 -4
bhl -38 -1 3 -29 -22 6 0 47 - 15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 .11
bo 1 -22 26 5 19 -13 23 45 -6 9 31 - 36
bo2 15 40-17 -44 43 4 -28 1 2 2 -30
w21 36 w32 _ o o 8 w04 2 -6 w4 2 19 w053 23 w5 3 19 w5 25 2 9
w22 -27 46 w33 -1 15 w04 3 12 w4 3 -44 w4 24 3 i-054 13 w54 49 w535 22
v23 14 -33 w34 46 42 w044 -47 w4 4 0 w4 34 16 w055 -20 w55 - , i 2 w54 5 -48
w24 18 -33 w35 39 - 7 w04 5 29 w4 5 43 w4 4 4 47 w 0 5 6 -38 w5 6 - 28 w5 5 5 2 2
w25 -18 10 w36 43 -21 w04 6 28 w4 6 -25 w4 54 22 w05 7 -13 w5 7 17 w 5 6 5 -21
w26 -6 9 w37 3 -Q w04 7 0 w4 7 14 w4 64 -42 wO 5 8 8 w58 5 w5 7 5 2 5
w2 7 -5 4 w3B 32 -14 w048 18 w4 8 -34 w4 74 5 w05 9 -24 w59 33 w585 -36
w28 22 32 w39 -20 32 wO49 -21 w4 9 17 w4 84 -36 w060 13 wCO 34 1.-596 -4 5
w29 39 -20 w40 39 -31 w050 -41 w50 -44 w4 95 30 w061 18 wCl - 20 w60C 11
w30 17 -46 w41 37 -19 v051 34 w51 0 w505 30 w06 2 O im i-6 2 24 w616 -43
w31 -40 -40 w05 2 22 w52 -49 wo 15 42
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain qu
reset run strain tall test
epoch 0 ss 2 7.561 ps s 2.7 851 gcor 0. 0000 cpn: ime P-2 in 20
ta 20 2 2 20 50 79 97 97 79 5 0 50 20 2 o 20 50 79 97 ; 57 79 50
ah 42 54 48 41 44 51 49 69 50 47 3 2 38 50 64 53 53 55 63 i 13 58 46
ao 46 52 46 67 52 52 60 42 45 65 4 7 62 70 31 34 59 57 39 : : 1 56 39
do -6- 13-12- 10 -1 6 7 14 8 -3 0- -10- 13 -6 n A, -2 4 12 : 11 5 o
bhl -38 - i 3 -29 -22 6 0 47 - -15 A - 25
bn2 -36 -9 -6 -15 6 27 44 - 26 35 -22
bo 1 -22 26 5 19 -13 23 4 5 -6 9 31 - 36
bo2 15 40 - 1 7 -44 43 4 -28 x n o -30
w21 36 X- 32 -22 8 xv-04 2 -6 x--4 2 O w053 2 3 T. T 1 O U 19 v:£ 2 5 2 9
w22 -27 4 6 w 33 _ i 15 x-04 3 12 w4 3 -44 w4 2 - o w054 13 w5 A 49 w 5 3 5 o n
w2 3 14 -33 w 34 46 42 w04 4 -47 w4 4 0 1-4 34 16 w055 -20 w55 - 12 w54 5 -48
w24 18 -33 w 3 5 39 - 7 w04 5 29 w4 5 43 w4 4 4 4 7 w056 -3S i- o 6 28 k 5 5 5 22
w2 5 -18 10 w 36 4 3 -21 w04 6 28 x.-4 6 -25 wA 5 4 22 w0 57 _ 1 0 1-0 l 17 iv* 5 6 5 -21
w26 -6 9 x.- 37 3 -9 w04 7 0 x-4 7 14 w4 64 -42 w058 8 w58 5 Tv 5 7 5 25
w27 - 5 4 w3S 32 -1 - 1-04 8 18 x-4 8 -3 4 k4 74 5 w0 59 -24 i-59 33 Tv* 5 85 -36
w28 22 32 w 39 -20 32 i-04 9 - 21 x-4 9 1 7 1.-484 -36 w06 0 13 w60 34 Tv 5 95 -45
w2 9 39 -20 w 40 3 S - 3 1 1.-05 0 1 i w5 0 -44 w4 9 5 30 1-061 18 w61 - 20 w6oe 11
w3 0 1 7 -4 6 x.- 4 1 w f _ i n i:0 51 .1 w -i w5 1 n V ' 5 05 30 -..-06 2 o i.-6 2 - 24 1.-S 16
w31 -40 -40 X.-0 5 2 2 2 i.-52 -49 15


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 0 tss 30.105 pss 2. 84 38 gcor 0. 0000 cpname pio ir i 50
ta 50 20 2 2 20 50 79 9 7 9 7 7 9 50 50 20 2 O 20 5C 79 S7 97 79
ah 4452484141-5149 69 50 4 7 32 38 50 64 5 3' 5 3 5 5 63 4 3 58 46
ao 45 52 46 G7 52 52 60 1 42 45 65 47 62 70 31 31 59 57 39 51 56 39
do 0 -8-12-14 -703 14 13 3 0 -3 -9 -G -6 -9 -2 8 11 9 9
bhl -38 -1 3 -29 -22 6 0 47 - 15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 m.n n
bo 1 -22 26 5 19 -13 23 45 -6 9 31 - 36
bo2 15 40 -17 -44 43 4 -28 1 2 2 -20
w21 36 w32 -22 8 1.-0 4 2 -G w4 2 19 :.-05 3 23 i.- 5 3 19 i.- 5 2 5 29
i.-22 -27 46 i.-33 -1 15 1.-0 13 12 i.-4 3 -44 i.-4 24 3 1.-05 4 1 3 -..-54 49 t.-5 3 5 O O
i.-23 14 -33 i.-34 16 4 2 1.-014 -4 7 i.-4 4 0 1,-4 34 16 w055 -20 i-5 5 - 12 l,* o 4 o -48
i.-24 18 -33 i.-35 39 -7 i.-04 5 29 w4 5 4 3 will 47 i.* 0 5 6 -38 i,-56 - 28 w5 5 5 o o
i.-25 -It 10 w3G 43 -21 I.-04G 28 wlG -25 wl 54 2 2 l," 0 5 7 -13 i.-5 7 17 1.-5G5 -21
w26 -6 9 w37 3 -9 1.-04 7 .0 i.-4 7 14 V.-4 6 4 -42 1.-058 8 i.-58 5 i.5 7 5 '1 T.
w27 -5 4 w3S 32 -14 w048 18 k4B -34 w4 7 4 5 wO 5 9 -24 V.-59 33 w5S5 -36
w28 22 32 i.-29 -20 32 1.-04 9 -21 i.-4 9 17 1.-4 84 -36 1.-060 13 w60 34 I.-596 -45
w29 -29 -20 w40 29 -21 v050 -41 i.-50 -44 k4 95 30 1.-0G1 18 t,-Gl - 20 1.-60G 11
v30 17 -46 i.-41 37 -19 w051 34 w51 0 vr50 5 30 1.-0G2 O 1.-62 21 w616 -43
w31 -40 -40 w052 22 i.-5 2 -49 i,-515 42
p to push/b to break/ to continue
disp/ exam/ get/ save/ set/ clea r cycle do log ne wstart F itrain qu
reset run strain tall test
epoch 0 e 10 L. S w u 322 pss 2 . 917 2 gcor 0. 0000 cpr.o .me p 11 i: , TO
ta 79 50 20 2 2 20 50 79 97 97 79 50 50 20 o 2 20 50 79 97 97
ah 47 50 47 41 44 51 49 69 50 4 7 32 38 50 64 53 * ^ < ^ W W k/ *J 63 43 58 46
ao 45 53 46 67 52 52 60 42 45 6 o 4 7 62 70 31 34 59 57 39 51 56 39
do 8 -1 -8 -14-11 -7 _ o 9 13 8 7 -3 -4 -2 -6- 14 -9 1 7 8 13
bhl -38 -1 3 -29 - 22 6 0 47 - 15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 3 5 _ 9 9
bo 1 -22 26 5 19 - 13 23 45 -6 9 31 - 36
bo2 15 40 -17 -44 43 4 -28 1 22 -30
w21 36 w32 -22 8 1,-04 2 -6 i,-4 2 19 i,-0 5 3 23 w53 19 i." 5 2 5 29
v:22 -27 46 w2 2 -1 15 1.-04 3 12 i.-4 3 -44 wl 24 3 w054 13 w54 4 9 w535 o o
w23 14 -33 i.-34 4 6 42 w044 -4 7 i,41 0 w4 34 16 w05 5 -20 i,5 5 -12 i.-54 5 -4£
w24 18 -33 i.-35 39 7 w045 29 wl 5 13 1.-4 4 4 4 7 w056 -38 i,-56 -28 i;5 5 5 2 2
i.-25 -18 10 i.-36 43 - 21 i,-04 6 28 w46 -25 w4 54 22 1.-057 -13 i.-5 7 17 w5 6 5 - 2 2
u-26 -6 9 i," 3 7 3 -9 1.-0 4 7 0 w4 7 14 1.-4 64 -4 2 wC 5 8 8 i,-5 8 5 w5 7 5 25
i,-27 -5 4 v38 32 - 1 4 1.04 8 18 i.-4 8 -34 1.-4 7 4 5 w059 O 1 ~ i. T wo 9 33 w585 -36
u-28 22 32 i.-39 -20 o n 1.-04 9 -21 wl 9 17 i,-4 84 -36 w060 13 1.-60 34 1.-596
v-2S 39 -20 v40 39 - 31 i,-050 -41 -.,-50 _ A A It 1.-4 95 30 wOC-1 18 wGl -20 wG 06
1.-30 17 -46 i.-4 : 3 7 - : 9 1.-0 51 3-1 i.-5 1 0 1.-505 30 1.-0C2 2 ... e 9 2 J w6 15 *i .'
:.*3 1 -10 -40 1.05 2 2 2 i.-5 2 -49 i.- 5 1 5 42


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tail test
epoch 0 ta 97 79 50 tss 36.251 20 2 2 20 pss 2 50 79 9 9283 7 97 gcor 0. 79 50 50 0000 20 cpnarce 2 2 20 pl 2 50 79 in 97 97
ah 4 9 4 o 4 7 41 44 51 49 G9 50 4 7 32 38 50 64 5 3 Z) 3 oo C 3 4 3 58 46
ao 45 53 4G C7 52 52 GO 42 45 G5 47 G2 70 31 34 59 57 39 51 56 35
do 13 6 0 -10-11-12 -9 2 9 8 12 0 1 bhl -38 -1 3-29 -22 G 0 47 - 15 4 - 25
bh2 -36 -9 -6 -15 6 27 4 4 -2G 35 -22
bo 1 -22 26 5 19-13 2 3 4 5 -G 9 31 - 3 G
bo2 15 40 -17-44 43 4 -28 1 2 2 -30
i>21 36 i.-32 -22 S w 04 2 -G i.-4 2 19 1.-053 23 wo 3 19 w 5 2 5 29
w3'2 -27 46 w33 -1 15 I.-0 4 2 12 i.-4 3 -4 4 i.-4 2 4 3 u-0 54 13 V 0 *i 4 9 **5 2 5 n r.
w2 3 14 -33 w34 4 C 4 2 wO4 1 -4 7 u-4 4 0 u-4 34 1C i>-055 -20 w5 5 - 1 O w 5 *i 5 -48
w24 1 O 0 lu w V w 3 5 29 7 w04 5 29 w4 5 43 w4 4 4 47 w05G -38 w 5 G - 28 t;5 5 5 .1 o
w25 -IS 10 w3G 43 -21 w04G 26 1-4 G -25 1.-4 54 2 w057 -13 k57 i 7 w5G 5 -21
k26 -6 9 w27 3 -9 t.-04 7 0 i>4 7 14 i.-4G4 -42 1.-058 8 k58 5 w5 7 5 25
w27 - 5 4 w38 32 -14 w C 4 8 18 w4 8 -34 w4 / 4 5 l.O 59 -24 In. fl 9 23 K l) 8 Zj -36
w28 22 32 i.-29 -20 3 2 1.-04 9 -21 i.-19 i - 1.-4 84 -3G 1.-060 13 kGO 1 o k5*-G -4 5
39 -20 v40 39 -31 w050 -41 i>5 0 -44 1--4 95 30 l-O G1 18 wGI - 20 In C 0 G 11
w30 17 -46 i.-41 37 -19 v051 34 i>-5 1 0 w50 5 30 kOG2 2 wG 2 24 kGI 6 -43
w3 1 -40 -40 <>052 22 i52 -49 i>-5 15 42
p to push/b to break/ to con tinue
disp/ exam/ get/ save/ set/ clear cycle do log nc -start p (train qu
rest t run strain taii test
epo ch '0 tss 38.893 pss 2.6426 f* or 0 . 0000 enname pl 2 1 J a 9 7
ta 97 97 79 50 20 2 2 20 50 79 97 o 7 7 9 50 50 20 2 2 20 50 79
ah 49 48 47 41 44 51 49 69 50 47 32 V C; 50 G4 53 53 55 S3 4 3 53 4G
ao 4 5 53 4 G 67 52 52 60 42 45 65 47 C 2 70 21 24 59 57 39 51 56 29
do 13 11 7 -4 -5-12-13 -5 i 4 13 c 1 4 4-10-12 -7 -7 -2 9
bhl -38 -1 3 -29 -22 6 0 4 7 - 15 4 - 2 5
bh2 -36 -9 -6-15 G 2 7 4 4 -2 6 35 _ n 2
bo 1 -22 26 5 19-13 23 45 -6 9 31 - 36
bo2 15 4 0 - 17 -44 43 4-28 1 2 2 -30
w21 36 w 32 e o g w04 2 -G 1--4 2 19 1.-053 23 i .* 5 2 n g 1.-5 2 5 29
w22 -27 4 6 i.- 33 -1 15 wO-13 12 w4 3 .1 J n ^ i-424 3 1-054 13 i:5 4 *; f 5 2 5 22
w23 14 -33 24 4 G 42 t."04 4 -4 7 w4 4 0 w4 34 16 w055 -20 w55 - u w5n 5 -4 8
w24 18 -33 w 35 39 -7 1.-04 5 2 5 w4 5 43 i.- 4 4 4 47 w056 -38 1.-5 6 - 28 u 5 5 5 22
w25 -18 10 i: 35 3 2i w04S 28 i.-4C -25 w 4 54 22 i.-0£7 -13 i." 5 7 17 Kj G 0 .01
w26 -6 9 W 27 3 -9 w04 7 0 i---l 7 1 4 w4 5 4 -42 w058 3 1.-58 5 5 7 5 o r
1-2 7 - 5 4 w 2 3 3 4 1.-04 8 18 1.-4 8 -34 w 4 7 4 5 w059 -24 i.-5 9 3 3 v Ci 8 c> -2G
i.-28 2 2 2 w 39 _OQ *5O w049 -21 w49 i 7 1.-4 84 -36 wOGO 13 wBC 3 j v5 96 - 4 5
29 39 -20 w 40 39 -31 w050 -41 woO -44 1-4 9 5 20 wOOl 18 wCl - 20 u-606 i i
w2 0 17 -46 4: 37-19 *.05 1 2- -.."5:1 Q 5 05 30 1.-0G2 2 wC 2 2-J vGi G - 3
-.:2 1 -40 -40 w0£2 22 i:52 -49 W "> i O j 9


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 0 tss 41.337 pss 2.4441 gcor 0. 0000 cpname pi 4 i n 79
la 79 97 97 79 50 20 2 2 20 50 79 97 97 79 50 50 20 2 2 20 50
ah 47 50 47 41 44 51 49 69 50 47 32 38 50 G4 53 53 55 63 43 58 46
ao 45 53 45 57 52 52 60 ' 42 45 65 47 62 70 31 34 59 57 39 51 5G 39
do 8 11 13 2 0 -8-12 :-10 -6-2 8 8 4 10 U -2 -8 -7- 12-9 2
bhl -38 -1 3 -29 -22 G 0 47-15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 -22
bo 1 -22 26 519 -13 23 45 -G 9 31- 3C
bo 2 15 40 -17 -44 43 4 -28 1 22 -30
t.*2 1 35 w32 -22 8 w042 -6 w42 19 w053 23 i.-5 3 19 w o 2 o 2 9
,.-2 2 -27 46 k33 -1 1 5 wOl 3 12 w4 3 14 wi 24 3 w05 1 13 w54 4 9 w5 3 5 O O
w23 14 -33 w34 46 42 w044 -47 w44 0 w434 16 w055 -20 wi 5 -12 w54 5 -48
*.*24 18 -33 k35 39 - 7 w0-!5 29 w4 5 4 3 w4 4 4 4 7 w056 -38 w5C -28 w5 5 5 .. >
w2 5 -18 10 w3G 43 -21 w04 6 28 w4 6 -25 w4 54 22 w057 -12 w 5 7 17 w5 6 5 -21
w2G -6 9 w37 3 -9 w047 0 w47 14 w464 -42 1.058 8 w58 5 w5 7 5 25
w27 -5 4 w38 32 -14 w048 18 w4e -34 w474 5 w059 -24 w59 -33 w58 5 -36
w28 22 32 w39 -20 3 2 w049 -21 w49 17 >.-484 -36 wOGO 13 wCO 34 w 596 4 o
i-29 39 -20 w40 39 -31 w050 -41 w50 -44 w495 30 w061 18 wG 1 -20 w60G 11
>3 0 17 -46 w41 37 -19 w051 34 w51 0 w505 30 w062 2 wG2 24 wG 16 -43
>3 1 -40 -40 w052 22 w52 -49 w515 42
p tc i push/b to brea)t/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do leg newstart ptrain qu
reset run strain tail test
epoc r: 0 tss 4 3.936 pss 2.5983 gccr 0. 0000 cpname pi 5 ii n 50
t a 50 79 97 97 79 50 20 2 2 20 50 79 97 97 7 9 50 50 20 2 2 20
ah 44 52 47 41 44 51 49 69 50 47 32 38 50 G4 53 53 55 63 43 5S 46
ao 45 52 46 67 52 52 60 42 45 65 47 62 70 31 34 59-57 39 51 56 39
do 1 7 13 6 6 0 -8 -10-11-10 0 4 4 14 10 -2 -1 -3- 12-12 -4
bhl -38 -1 3 -29 -22 6 0 47-15 4- 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 -22
bci -22 26 5 19 -13 23 45 -G 9 31- 36
bo 2 15 40 -17 -44 43 4 -28 1 22 -30
-21 36 w32 -22 8 w04 2 -6 w42 19 w053 23 w53 19 wo 2 5 29
:*/>2 -27 46 w33 -1 1 5 w04 3 12 w4 3 -44 w424 3 w054 13 w54 4 9 w5 3 5 22
w23 14 -33 w34 46 42 w 0 4 n 4 i i." 4 4 0 w : 3 4 1 6 w055 -20 w55 -12 w5 4 5 -48
w24 18 -33 v35 39 7 w045 29 w45 43 w444 4 T w056 -38 i.-56 -28 VT l> 0 0 22
w2 5 -18 10 w36 43 - 21 w046 28 w46 -25 w454 a o w057 -13 w5 7 17 wo 6 5 - 21
w26 -6 9 w37 3 . -9 w04 7 0 w47 14 w4 64 -42 wO 58 8 wo 8 5 wo 7 5 O n
-27 -5 4 w3S 32 -1 4 w048 18 w48 -34 w474 5 w059 -24 w59 33 wo 8 5 -36
<-28 22 32 i.-39 -20 32 w049 -21 w49 17 w484 -36 w050 13 w60 3 4 w596 -45
-29 39 -20 w40 39 -31 w050 -41 w50 -44 w495 30 wOGl IS wCl -20 w6 0 6
..3 0 17 -45 w41 37 -19 w0 51 34 w5l 0 1.-505 30 i.-062 2 wGE 24 wcie __ i ? 1 w
*' 31 -40 -40 w052 2 2 wi -49 w515 42


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 0 tss 4G.714 pss 2 . 7787 gcor 0. 0000 cpname plG in 20
ta 20' 50 79 97 97 79 50 20 o 2 20 50 79 97 97 7 9 50 50 20 2 2
ah 42 54 47 41 44 51 49 69 50 4 7 32 38 50 G4 53 5 3 5 5 63 43 58 4 G
ao 4G 52 4G G7 52 52 60 42 45 C5 47 G2 70 31 34 59 57 39 51 5 G 39
do -G 0 8 G 10 G -1 -G- 11-14 -G -2 0 14 14 4 -2 3 -7-12 -8
bhl -38 -1 3 -29 -22 6 0 4 i 1 a 4 - 25
bh2 -36 -9 -G -15 6 27 44 -2G 35 -22
bo 1 -22 2G 5 19-13 23 45 -G 9 31 - 3G
bo2 15 40 -17 -44 43 4 -28 1 22 -30
v21 36 w32 -22 8 1.-04 2 -G w4 2 19 1.-05 3 23 i.-5 3 19 w525 29
w22 -27 4G w33 -1 15 wO-13 12 i.-4 3 -44 4 24 O V 1.-054 13 w54 49 w 5 3 5 r r,
w2 3 14 -33 w34 4 G 4 2 w04 4 -47 i.'4 4 0 w4 3 4 16 1.-0 55 -20 i.*5 5 - 12 w54 5 -48
w24 18 -33 w3 5 3 9 -7 w04 5 29 i.-4 5 4 3 w4 4 4 47 1.-05G -38 w5G - 28 i.- 5 5 5 22
w25 -18 10 w36 43 -21 w04G 28 w4C -25 w4 54 22 w05 7 - 1 3 w 5 7 1 7 w 5 G 5 -21
w2G -6 9 w 3 ? 3 9 w04 7 0 i.-4 7 14 w4 6 4 -42 w058 8 i.-58 5 i."5 7 5 25
w2 7 -5 4 w38 32 -14 w04 B 18 t.-46 -34 w4 7 4 5 w059 -24 w59 3 3 w 5 8 5 -3G
w28 22 32 w39 -20 32 w0 4 9 -21 w4 9 17 w4 84 -36 1.-060 13 v-60 34 w59G -4 5
i.-29 39 -20 v40 39 -31 w050 -41 w 5 0 4 4 1.-4 95 30 wOGl 18 i.-61 - 20 1.-60G 11
w30 17 -4G w41 37-19 u-05 1 34 w51 0 w505 30 w062 2 wG2 24 wG 16 -43
w31 -40 -40 u-052 22 w52 -49 w515 42
p to push/b to breai:/ to continue:
c:sp/ e::am/ get/ save/ set/ clear cycle do log newstart ptrain qui
reset run strain tall test
epoch 0 tss 49. 54 4 pss 2 . 8297 gcor 0. 0000 cpname Pi 7 4 rj O
ta 2 20 5 0 79 97 97 79 50 20 2 2 20 50 7 9 97 97 79 50 50 20 2
ah 40 55 4 1 41 4 4 01 49 G9 50 4 7 32 38 50 64 53 53 55 G 3 4 3 58 4 6
ao 46 52 46 G7 52 52 GO 42 4 5 G 5 4 7 G 2 70 31 34 59 57 39 51 5 G 3 9
do -10 7 1 2 9 11 <1 1 -7-1 5-12-10 -5 10 13 9 5 0 0 -7 -8
bhl -38 _ i 3 -29 - 22 G 0 4 7 -1 5 4 - 25
bh2 -36 -9 -6 -15 6 0 T -1 A 1 n -26 35 0 0 t
bol -22 26 5 19 - 13 23 4 5 -6 9 31 - 36
bo 2 1 5 40 -17 -44 43 4 -28 1 22 -30
*.."21 36 ..OO .OO 0 w04 2 -G '-4 2 19 w05 3 23 wo 3 19 w 5 2 5 29
u-22 -27 46 w33 -1 1 5 w04 3 12 w4 3, -4 4 v4 2 4 3 w054 13 wo 4 49 vo 35 22
t.-2 3 i 4 - 3 3 w34 46 42 1,-04 4 -4 7 w4 4 0 I.-4 34 1G w05 5 -20 w5 5 - 1 2 i.-5 4 5 -48
w24 18 -33 i.-3 5 39 - 7 w04 5 29 w4 5 4 3 w4 4 4 47 w056 -38 v55 - 28 w55 5 2 2
i.-25 -18 10 w3G 43 - 21 w046 28 w46 -25 w 4 0 4 0 2 i.'0 5 7 -13 wo 7 1 1 w5 6 5 -21
v26 -6 9 w37 3 -9 u-04 7 0 i.-4 7 14 1.-4 64 _ j 0 I.-05S e i.-58 5 wo 7 5 25
w2 7 - 5 1 7 w38 32 - 14 1.-04 8 18 w48 -34 i.-4 74 5 w059 -24 vo 9 33 wo 8 5 -36
w28 22 32 i.-39 -20 32 I.-04 9 -21 w4 9 17 wie4 - 3 G w060 1 3 w60 34 wo 96 -4 5
i.-29 29 -20 w40 39 - 3 1 I.-050 -41 w50 -44 1.-4 95 30 wOGl 18 w61 - 20 w6 06 1 i
v-30 17 -46 i-4i 2 7 - 19 ..-0 51 3 1 t.-51 0 i.-505 30 w062 0 t.-62 24 i-:£ 2 G -43
w31 -40 -40 0 ~ r> 0 C* w52 -4 9 1.-515 - 2


quit
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain
reset run strain tall test
epocn 0 tss 52.492 pss 2.9483 gcor 0.0000 cpname p!8 in
ta 2 2 20 50 79 97 97 79 50 20 2 2 20 50 79 97 97 79 50 50 20
ah 40 55 48 41 44 51 49 09 50 47 32 38 50 G4 53 53 55 63 4 3 58 46
ao 4C 52 46 C7 52 52 60 42 45 65 ; 47 62 7 0 31 3 4' 5 9 5 7 39 51 5 G 39
do -11-12 -6 -3 5 11 8 9 0-11 -12-14- 10 4 9 9 9 9 0 -1 -4
bh 1 -38 -1 3 -29 -22 6 0 4 7 -15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 - 26 25 _ } o
bol -22 20 5 19 -13 23 45 -G 9 31 - 36
bo2 15 40-17 -44 43 4 -28 t no -30
w21 36 w32 _ n o 8 t>04 2 -6 w 4 2 19 w053 23 K 0 3 19 u-525 29
w22 -27 46 w33 -1 15 w04 3 12 w43 -4 4 w-4 24 3 w054 1 3 w 54 49 tv 5 3 5 *v o
1.-23 14 -23 w34 4 6 42 v-044 -47 w 4 4 0 w4 34 16 w0 5 5 -20 w5 5 - 12 wo 4 o -4e
. 4 V 4. ~l 18 -33 w25 3 9 -7 w04 5 29 i.' 4 o i 3 t. 4 4 4 4 7 w056 -38 5 G - 28 tv-5 5 5 o o
w25 -18 10 w36 43 -21 w04 6 28 w46 -25 w4 54 A t> wO 5 7 -13 5 7 17 w5G 5 - 21
w2G -6 9 w37 3 -9 w04 7 0 w 4 7 14 w 4 6 4 A w-2 7 -5 4 w38 32 -14 v048 18 w 4 6 2 4 w4 7 4 5 w059 -24 w59 33 w 5 8 5 -36
w28 22 32 w-39 -20 3 2 w-04 9 -21 w-19 17 w 4 8 4 -36 w060 13 w60 34 tv-5 96 - 4 5
w29 39 -20 w-40 39 -31 w050 -41 w o 0 4 4 w4 95 30 t.-061 IS kCI - 20 w60G i
w-30 17 -46 w41 37 -19 w051 34 wol 0 w-505 30 k06 2 o wG2 24 tv-616 -4 3
31 -40 -40 v-052 22 w52 -49 wo 15 42
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ ciea .r cycle do log newstar ptrain quit
reset run strain tail test
epoch 0 tss 55.509 pss 3.0163 gcor 0. 0000 cpn: ame p!9 r. 20
ta 20 A 2 20 50 79 97 97 79 50 20 2 2 20 50 79 97 97 79 50 50
ah 42 54 4 S 41 44 51 49 69 50 47 32 38 50 64 53 53 55 63 43 58 46
ao 46 52 46 67 52 52 60 42 45 65 47 62 70 31 34 59 57 39 51 56 39
cc -6- 13-1 2-10 -167 14 8 -4 -7- 14-13 -2 3 5 8 13 7-1 n
bh 1 -38 -1 3 -29 -22 6 0 4? - 15 4 - 25
bh2 -36 -9 -6 -15 6 27 44 -26 35 -22
bo 1 -22 26 5 19 -13 23 45 -6 9 31 - 36
bo 2 15 40 -17 -44 43 4 -28 1 22 -30
w21 36 tv-32 -22 8 i.-04 2 0 tv* 4 2 19 tv-05 3 2 3 t:5 3 19 w525 L 9
tv*0 2 -27 46 v23 -1 15 t.-04.3 12 u-4 3 -44 1--4 24 3 w0 54 13 w-54 4 9 tv 5 3 5 A A
w23 14 -33 tv-34 46 42 w-04 4 -4 7 tv4 4 0 w434 16 w05t -20 t.-5 5 -12 w-54 5 -48
i.-2 4 18 -33 u-35 39 -7 1.-C4 5 29 k4 5 43 w-444 4 7 w056 -38 tv-56 -28 w555 22
w25 -18 10 w-36 43 -21 w-04 6 28 tv-4 6 -25 w454 2 2 v 0 5 7 -13 w-5 7 17 t.-5C5 -21
tv-2 6 -6 9 u-3 i 3 -9 w04 7 0 tv-4 7 14 t,-4 6 4 -42 t-0 58 8 1.-58 5 w-5 7 5 25
v 2 7 - 5 A T u-38 32 -14 tv-04 8 18 tv-4 8 - 3 4 I.- 4 7 4 5 t,-0 59 -24 w-5 9 n r> tv- 5 8 5 -36
w28 22 32 i-*39 -20 32 tv-04 9 -2! t.-4 9 17 w484 -36 u-060 13 w-eo 34 t.- 5 9 6 0
1--2 9 o O -20 k40 39 -21 wC5 0 -4 1 t.-5 0 -44 w495 30 w061 18 V61 -20 k606 1 1
w 3 0 17 1 C n t-- i*4 1 37 1 9 tv-r-f-i 84 0 t.- 50? 3 0 tv-0 6 2 2 w-6 2 24 wC : C --I3
I--31 -40 -40 w05 2 22 I.-5 2 -4 9 t.-515 42
a c i co ci


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ dear cycle do log newstart ptrain quit
reset run strain tall test
epoch 0 tss 58. 448 pss 2 . 9394 gcor 0. 0000 cpname p20 in 50
ta 50 20 2 O 20 50 79 97 97 7 9 50 20 2 2 20 50 79 97 97 79 50
ah 44 52 48 41 44 51 49 69 50 4 7 32 38 50 64 53 5 3 5 5 63 4 3 58 46
ao 4 5 52 46 67 52 52 60 42 45 65 47 62 70 2 1 34 5 9 5 7 39 51 56 39
do 0 -8-12 -14 -7 0 3 14 13 3 0- -10- 13 -6 _ o -2 4 12 11 5 o
bhl -38 -1 3 -29 - 22 6 0 47 - 1 5 4 - 25
bh2 -36 -9 -6 - 1 5 6 27 44 -26 35 _ o o
bol -22 26 5 19 - 13 23 ^ r -6 9 31 - 36
bo 2 15 40 -17 -44 43 4 -28 1 22 -30
w21 36 w32 8 w04 2 -6 w4 2 19 u-053 23 i.*5 3 19 u-525 29
w22 -27 46 w33 _ 1 15 w 0.3 12 w4 3 -44 w4 24 3 w054 1 3 i.-54 49 w 5 3 5
i.-23 14 -33 w34 46 42 w04 4 -4 7 w4 4 0 i.-4 34 16 5 5 -20 w5 5 t 12 w54 5 -48
i.-24 18 -33 w35 39 - 7 w0 4 5 29 w4 5 43 w4 4 4 47 i> 0 5 6 -38 i." 5 6 - 28 i." 5 5 5 *1 O
w2 5 -18 10 w36 43 - 21 w04 6 28 u-4 6 -25 w4 54 22 1.-05 7 -13 w5 7 1 7 w 5 6 5 -21
k26 -6 9 w3 7 3 -9 w04 7 C w4 7 14 w 4 6 4 -42 w0 58 8 w58 5 i.- 5 7 5 25
w27 -5 4 w38 32 - 14 w04 8 18 w4 8 -34 w4 7 4 5 i.-05 9 -24 i.-5 9 33 1. 0 8 3 -3G
w28 22 32 w39 -20 32 w04 9 -21 w4 9 17 w 4 8 4 -3G w0 60 13 1.-60 34 w59 6 -4 5
w29 39 -20 w4 0 39 - 31 w 0 5 0 -4 1 w50 -44 i.-4 9 5 30 u-061 IS i.-e i - 20 l.-GOG 11
w30 17 -46 w4 1 37 - 19 w051 34 w51 0 w505 30 w062 2 w62 24 i.-61G -43
w21 -40 -40 wO 5 2 22 w52 -49 w515 42


n-sair.pl e delay network displays
sinusoidal
input patterns p.G-p2C
or
p to push/b to break/ to continue:
disp/ exam/ set/ save/ set/ clear cycle do log newstart ptrain qui
reset run strain tall test
epoch 247 tss 5.0093 pss 5.0093 ecor 0. 0000 cpname p. 0 in 50
ta 50 50 20 2 2 20 50 79 97. 97 79 50 20 2 0 2 0 5 0 79 97 97 79
ah 44 50 35 35 29 44 34 44 20 2B 42 31 38 31 2 3 10 14 99 4 9 8
ao 50 7056352 1/ 4 o 5 0 7 99 99 0 9C 34
do 03000022 5 4 3 2 0 0 0 0 0 0 0 0 10
bhl -41 -10 -01 -59 -86 -21 -66 -23-135 -91 - 31
bh2 -77 -47 -77-]16-218-1 7G 491-310-228-223
bo 1-200-218-25C-220-239-195- 221-200-249-232-199
bo2-227-205-221-236-248 706 127-275 225 -14
v2i 37 >.-32-150-199 w042 54] v42 -21 >.053 402 >53-204 w 5 2 5 22
>.-22 75 -90 >.-31-190- 163 >.-042 424 >.-43- 125 >.424 80 >.054 3 2 2 v 5 4 2 2 £ w 52 5 9
>.-23 02 -32 >.-34- 1 8G- 100 w044 4 4 j w4 i *i>0 ii Ji 59 >. 0 5 5 407 >.-55-192 v5 4 5 10
w24-129-158 w35- 1 76 -7 1 w045 4 0 d w 4 J) 2 / C w44 4 34 v0 5G 4 2 5 W5G-158 w 5 5 5 18
w25 -39-1 10 v-30-1 07 -05 >.-040 415 w4G-101 w4 54 50 w 0 5 7 4 0? k 5 7 7 3 v5 G 5 00
>.-20-298-240 w37-202-155 w047 3/9 w 4 / J fc 4 w 4 G 4 18 w056- 4 19 w 5 8 8 9 w 5 7 5 -247
>.-2 7-210-150 u-36-580-38.2 w048 .401 w4 8-201 w4 7 4 27 w 059- 172 w5S-417 w585 1042
>.-28-245-274 w39-281-281 >.-049 270 >.-4 9-27 0 v-4 84 19 wOOO 1140 >.00- 1 01 >5 90 -910
W29-124 -91 U-40-1C9-125 wOOO 421 u-50-134 w4 95 36 >.001- 4 55 wC] C wOOO -209
w30-1 65-149 u-41-240-13 1 w051 4 03 v51- 185 w505 29 wOG2 85 u-02 -4 4 V.-616 -45
u-31-216-171 w052 387 w52-24 1 >.'515 3
p to push/b to breal;/ to continue:
cisp/ exam/ set/ save/ set/ clear cycle do log newstart ptrair. qui
reset rur. strain tall te s t
epoch 247 tss 11.370 pss 6.3071 gcor 0. 0000 cpr.ame P-1 i: r. 7 9
ta 79 50 50 20 2 2 20 50 79 97 97 79 5C 20 2 2 20 50 79 97 97
ah 47 51 41 31 27 36 29 38 18 25 30 27 34 28 21 9 11 6 5 0 1 2
ao 82 48 10 7 e 6 6 5 7 7 C 0 G 7 1 5 99 0 9 18 94 55
do 00200002 4 C 5 4 2 C 0 0 0 0 9 0 10
bkl -41 -1C -51 -59 -80 -21 -60 -23-135 -91 - 21
bh2 -77 -47 -77-118-218-170 491-310-228-233
bo1-260-218-250-220-229-195- £21-200-249-232-195
be2-227-205-222-238-248 766 127-375 .225 -14
v2i 37 v32-i 56-199 >.'042 541 >.-4 2 -21 v05 3 402 1.-53-2C4 W O 2 Li 0 0
-22 75 -90 w33-190-103 >.'043 424 v-4 2- 125 w424 60 w054 *) 0 w54-22c w 5 2 5 9
>.'23 02 -32 >.-34-186-100 v044 443 v* 00 v*1 o4 59 w055 407 W55-192 w5 4 0 16
v24-129-156 >.35-176 -71 >.-045 4 05 v-4 5-1 76 w4 4 4 24 >.0 56 4 29 w 0 u 2tS U D d J 1 O >_
v25 -39-110 v:36 -107 -05 >.-040 4 ; w 4 G -1 0 : v 4 o 4 50 vG5 7 403 v 5 7 7 3 v56 5 60
-26-298-240 w37-202-155 >.-047 279 >.-4 7-284 v464 ie wC5S- 419 >58 -69 V L 1 Z) -24 7
-27-210-150 W38-566-382 >.-048 4 C 1 >.-46-201 v 4 7 4 27 w059- 172 w 19 4 2 ; w 5 £ 5 1042
-2S-245-274 w39-2£1-281 k049 276 w49-276 w484 19 v-0001 140 vCO-101 w5 50 -910
>.29-124 -91 >.-40-169-125 >.-050 421 >.-50- 1 24 -..-4 95 2 £ >.061- -1 1/ c wCl 0
-30-165-149 w4 1-24C-1 31 wOol 4 02 w 5 1 1 8 C w 5 0 5 29 >062 £5 >.-C2 -44 >.016 nil
v-31-216-171 ;.-052 287 w52-241 v&l5 0


quit
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain
reset run strain tall test
epoch 247 tss 16.804 pss 5.4284 gcor 0.0000 cpname P' , 9 in 9 7
ta 97 79 50 50 20 2 2 20 50 79 97 97 79 50 20 2 2 20 50 79 97
ah 48 46 44 19 26 33 27 35 18 24 35 26 32 27 20 8 10 52 0 1 1
ao 92 84 47 12 9 7 7 6 7 8 7 7 7 8 8 10 99 99 42 81 59
do 0 0 0 4 0 0 0 0 3 5 6 6 5 3 0 0 0 0 o -1 9
bhl -41 -10 -61 -59 -86 - 21 -66 -23- 135 -91 -31
bh2 -77 -47 -77-11B-218-176 491-316-228-233
bo 1-260-218-256-220-239-195-221-200-249-232-199
bo2-227-205- 222-238-248 766 127- 375 325 -14
w21 37 u-32-156-199 wO 4 2 54 1 w42 -21 w05 3 402 1.-53-204 w525 22
w22 75 -90 w33-190-16 3 w04 3 424 1.-4 3-125 w4 24 80 w054 392 u-54-228 w5 3 5 9
k23 62 -32 W34-186-100 w04 4 443 w44 -50 w4 3 4 59 w05 5 407 w55-192 w54 5 16
w24-129-158 w35-176 -71 wO 4 5 405 W45-176 w4 4 4 34 u-056 429 k5G- 1 58 w55 5 18
w25 -39-110 w36-107 -65 w04 6 415 w46-101 w454 50 w057 403 u-57 -73 k5G5 60
v.26-298-246 i.-37-202-1 55 w04 7 379 w47-284 w4 64 18 w058- 419 w58 -89 w575 -247
w27-210-150 W38-586-382 w0 4 8 401 w48-201 w4 74 27 w059- 172 w59-417 w585 1042
w28-24 5-274 w39-281-281 w04 9 376 w4 9-276 w4 84 19 W0601140 w60-101 w59G -910
W29-124 -91 w40-l69-125 w050 421 w50-134 w4 95 36 wOG 1 - 455 w61 0 wGOG -209
w30-165-149 w41-24 0-131 w05 1 403 w5 1-185 w50 5 29 w06 2 85 w62 -44 w616 -4 5
w31-216-171 w052 387 w52-241 w515 3
p to push/b to break/ to continue:
disp/ exam/ get/ save/ si et/ clear i cycle do log nei. start ] ptrain ui
reset run strain tall test
epoch 247 tss 21.356 pss 4.5518 gcor 0. 0000 cpna .me p. 3 ii n 9 7
ta 97 97 79 50 50 20 2 2 20 50 79 97 97 79 50 20 2 o 20 50 79
ah 48 43 43 7 22 29 26 34 17 23 34 25 31 26 19 8 10 50 0 0 0
ao 92 91 84 48 13 8 8 6 8 8 7 7 7 8 9 10 99 99 44 57 59
do 00004000 13 5 6 6 5 3 0 0 0 -4 -2 4
bhl -41 -10 -Cl -59 -86 -21 -66 -23-135 -91 - 31
bh2 -77 -47 -77-118-218-176 491-316-228- 233
bol-260-218-256-220-239-195- 221-200-249- 232-199
bo2-227-205-222-238-248 766 127-375 325 -14
w21 37 w32-156-199 w042 541 v42 -21 u-053 402 1.-53-204 w525 22
w22 75 -90 w33-190-163 w043 424 w43-125 u-4 24 80 1-054 392 1.-54-228 i-5 35 9
w23 62 -32 w34-186-100 w044 443 w4i -50 w4 34 59 w05 5 407 i.- 5 5 -19 2 w54 5 16
w24-129-158 w35-176 -71 w045 405 w45-176 w4 4 4 34 w056 429 1.-56-158 w555 18
w25 -39-110 I.-36-107 -65 w04G 415 I.-46-101 u-4 54 50 w057 402 i.-57 -72 w 5 6 5 60
w26-298-246 w37-202-155 w047 379 u-4 7-284 u-4 64 18 i.-058- 418 -..-58 -89 w575 - 24 7
u-27-210-150 w38-58C-382 w048 401 1.-48-201 u-4 74 27 v059- 172 u-59-4 1 7 i.- 5 8 5 1042
w28-245-274 w39-231-281 w049 376 u-49-276 w4 84 19 1.-0601 140 1.-60-101 V.-596 -910
w29-124 -91 -..-40-169-125 -..-050 421 i." 5 0 -13 4 1.-4 9 5 36 w061- 455 w51 0 w606 -209
--30-155-14 9 wi 1-240-131 -.-05: 403 w51-185 1.-505 29 i.-062 85 i.-62 -44 1.616 -4 5
w31-216-171 w052 387 1.-52-241 wo 15 2


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 247 tss 25.462 pss 4 . 1056 gcor 0.0000 cpname P' , 4 i n
ta 79 97 97 79 50 50 20 o 2 20 50 79 97 97 79 50 20 2 2 20 50
ah 47 39 42 3 14 11 24 32 17 23 33 25 31 26 19 8 10 49 0 0 1
ao 82 91 90 84 49 16 9 7 8 8 7 8 8 8 9 11 99 99 45 51 54
do 0 0 0 0 0 4 0 0 0 0 3 5 6 7 6 O 0 0 -6 - 7 _ i
bhl -41 - 10 - 61 -59 -86 - 21 -66 -23- 135 -91 -31
bh2 -77 -47 -77-113-218-176 491-31G-228-233
bol-260-218-256-220-239-195-221-200-249-232-199
bo2-227-205- -222-238-248 766 127- 375 325 - -14
w21 3 7 >-32-156-199 >-042 54 1 >-4 2 -21 v053 402 >-53-204 >-525 22
w22 75 -90 >-33-190-163 >-04 3 424 w-13-125 w4 24 80 >-054 392 w54-228 w535 g
w23 62 -32 w34-186-100 w044 443 w44 -50 w4 34 59 w055 407 >-55-192 w54 5 1G
>-24-129-158 >-35-1 76 -71 >-04 5 405 >-4 5-176 w4 4 4 34 w056 429 w a 6 1 o 8 w5 55 18
w25 -39-110 >-36-107 -65 w04 6 415 W4G-101 w4 54 50 w057 4 03 >-57 7 3 w56 5 60
>-26-298-246 k37-202-155 w04 7 379 w4 7-284 w4 64 18 w058- 419 >-58 -89 w5 7 5 -247
>-27-210-150 w38-586-382 w04 8 401 W48-201 w4 7 4 27 w059- 172 w 5 9-4 17 w585 1042
w'28-245-274 w39-281-281 w04 9 3 7 6 w4 9-276 w4 84 19 W0601140 >-60-101 >-596 -910
w29-124 -91 w4 0-169-125 w050 421 >-50-134 w4 95 36 >-'0G 1 - A 55 w61 0 >-606 -209
w30-165-14 9 v41-24 0-131 V 0 D 1 403 k51-185 w505 29 >-062 85 wG2 -44 >-616 -45
31-216-17 1 v052 387 w52-241 w515 3
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 247 tss 29.179 pss 3. 7172 gcor 0. 0000 cpr*ame p. 5 in 50
ta 50 79 97 97 79 50 50 20 2 2 20 50 79 97 97 79 50 20 2 2 20
ah 44 36 40 3 9 1 12 28 17 23 33 24 31 26 19 8 10 48 0 0 1
ao 50 84 90 88 84 51 16 8 8 8 8 8 8 8 9 11 99 99 45 50 52
do 00000040 0 0 0 3 5 7 7 6 0 0 -4-11 -8
bhl -41 -10 -61 -59 -86 -21 -66 -23-135 -91 - 31
bh2 -77 -47 -77-118-218-176 491- 316-228- 233
bol-260-218-256-220-239-195- 2 21- 200-24S- 232-199
bo2-227-205-222-238-248 766 127- 375 325 -14
v21 37 w32-156-199 w042 541 >.-4 2 -21 w053 402 >.53-204 >.-525 2 2
v22 75 -90 w33-190-163 v043 424 >4 3-125 w4 24 80 w054 392 >.-54-228 >."5 3 5 9
>,-23 62 -32 w34-186-100 >.-044 443 w44 -50 w434 59 w05 5 407 >.55-192 w54 5 16
v-24-129-153 w35-176 -71 w045 405 >.4 5-176 w4 4 4 34 w056 429 w56-l58 v5 55 18
w25 -39-110 >.-36-107 -65 w046 4 15 >.-4G-101 w4 o4 50 >,'057 403 w 3 t i 3 >.56 5 60
U-2G-298-246 w37-202-155 w047 379 v4 7-284 >.4 64 18 w058- 4 19 w58 -89 w 0/0 -247
u-27-210-150 >.-38-586-382 w048 401 >,48-201 >.4 7 4 27 w059- 172 >.-59-417 >.585 104 2
w28-245-274 w39-261-281 >.-049 376 >.49-276 >.-4 84 19 >.0601 140 u-60-101 >596 -910
>,29-124 -91 w40-169- 125 w050 421 >-50-134 >. 4 9 o 36 wOGl- 4 5 5 wGl 0 w60G -209
>30-165-149 >.-41-240-1 21 >.051 403 >51-125 >.- 5 0 5 29 >062 85 >.-G2 -4 4 w516 -45
-31-216-171 >.-052 387 >52-24 1 >.*515 rj -


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 247 tss 32.332 pss 3 .1526 gcor 0.0000 cpname P- 6 i n
ta 20 50 79 97 97 79 50 50 20 o o 20 50 79 97 97 79 50 20 2 2
ah 41 38 36 3 8 0 3 11 15 22 33 24 3 1 26 19 8 10 48 0 0 1
ao 17 49 es 89 88 82 52 17 9 8 8 8 8 8 9 11 99 99 45 50 52
do 0 0 0 0 0 0 0 4 0 0 0 0 3 5 7 8 0 0 0- 10- 12
bhl -4 1 -10 -Cl - 5 9 -8G - 21 -GG -23- 135 -91 -31
bh2 -77 -4 7 -7 7-118-218-17G -19 1 -31G-228-223
bo1-2G0-218-256-220-239-195-22]-200-24 9-232-199
bo2-227-205- 222-238-248 7CC 1 27- 375 325 -14
w21 37 1.-32-15C-199 w04 2 541 i:i 2 -21 i.053 402 t:53-204 i:525 n *}
w22 75 -90 w3 3-190-163 i:04 3 424 i:4 3-125 w4 24 80 i:0 5 4 392 i:54 -228 i:535 9
w23 G2 -32 1.-34-18G-100 w04 4 4 4 3 w44 -50 i:4 84 59 w055 4 07 i:5 5-192 w o 4 a 1G
w24-129-158 w3 5-17 6 -7 1 w04 5 405 u-45-176 w4 4 4. 34 i:05C 429 i:56- 1 58 i:5 5 5 18
w25 -39-110 w36-107 -65 w04 6 415 w4 6-101 w4 54 50 w05 7 4 03 w 5 7 7 3 w565 60
w26-298-246 w3 7-202-15 5 i:04 7 379 w47-284 i:4 6 4 18 w058- 4 19 i:58 -89 w5 7 5 -247
w27-210-150 w38-586-382 w04 8 401 W48-201 i.-4 7 4 27 w059- 172 w59-4 17 v585 1042
i:28-245-274 w39-281-281 w04 9 376 i:4 9-27-6 w 4 8 4 19 i:060114 0 1.-60-101 i: 5 9 C -910
W29-124 -91 w4 0-16 9-125 w050 421 woO-134 w4 9 5 36 w06 1 - 4 5 5 i:61 0 i.-GOG -209
w3 0-165-149 w41-24 0-131 w051 403 w51-185 w50 5 29 i:0G2 85 i:62 -44 i:G 1 G -4 5
w31-21G-171 w052 387 W52-241 w51 5 3
p to push/b to break/ to continue:
disp/ exam/ get/ save/ se t/ clear < cycle do log nei. start p (train quit
reset run strain tall test
epoch 247 tss 34.698 pss 2.3667 geer 0. 0000 cpname p. 7 ir r>
ta 2 20 50 79 97 97 79 50 50 20 2 2 20 50 79 9 7 9 7 79 50 20 2
ah 40 42 23 5 8 0 1 1 9 20 33 24 31 26 IS 8 10 48 0 0 1
ao 7 15 49 85 88 86 82 53 16 9 8 8 8 8 9 11 99 c 9 45 50 52
do 0 0 000 1 004000 0 3 6 8 0 0 4 -6-12
bhl -41 -10 -61 -59 -86 -21 -65 -23-135 -91 - 31
bh2 -77 -47 -77-118-218-176 491-316-228- 233
bo1-260-218-256-220-239-195-221-200-249- 232-199
bo2-227-205-222-238-248 766 127-375 325 w21 37 w-32-156-199 w042 541 i.-4 2 -21 -14- w053 402 '5 3-204 '.-525 22
w22 75 -90 W33-190-163 i:043 424 w43-125 w4 24 80 i:0 5 4 392 W54-228 i:5 3 5 a
i:23 62 -32 1:34-186-100 w044 443 w44 -50 ': -i J 4 59 w055 407 wo5-192 w5 4 5 16
w24-129-158 w35-17C -71 i:045 405 i:45-176 w44 4 34 w056 4 29 w 5 G 15 8 w 5 5 5 18
i:2 5 -39-110 i:36-107 -65 i:045 4 15 i: 4 C -101 w4 5 4 50 i:05 7 403 w 5 7 7 3 w 5 G 5 CO
w26-298-246 i:37-202-155 v047 379 w47-284 v4 64 18 t:058- 419 w58 -89 ':575 rt \ 7 U i
-27-210-150 w38-586-382 w048 401 w48-201 k4 7 4 27 w059- 172 '59-417 vo 8 5 1042
'-28-245-274 i:39-281-281 w049 376 u-49-276 '4 84 19 '.0601 140 '.60-101 wo 9 6 -9:0
'29- 124 -91 '.-40-169- 125 w050 421 '.-50-134 w4 95 36 '061- 455 '61 0 i:6 0G -209
'30-165-149 v41-240-131 '.-051 403 i:51-!85 i:5 05 9 .:06 2 3 5 .:C 2 -4-1 'GIG - 4 5
-..31-216-171 '.052 387 w52-241 '.-515 3


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 247 tss 36.330 jjss 1.6320 gcor 0. 0000 cpname P< ,8 in
ta o 2 20 50 79 97 97 79 50- 50 20 o 2 20 50 79 97 97 79 50 20
ah 40 45 34 14 10 0 1 0 4 10.29 24 30 26 19 8 10 43 0 0 1
ao 7 9 14 49 84 87 86 81 51 15 9 8 8 8 9 11 99 99 4 5 50 52
do 0 0 0 0 0 1 1 0 0 4 0 0 0 1 3 6 0 0 8 0 7
bhl -4 1 -10 -61 -59 -86 -21 -66 -23- 135 -91 - 31
bh2 -77 - 4 7 -77- 118-218- 176 491-316- 228- 9 O C
bo 1 -260 -21S-256- 220-239- 195- 221-200- 249- 9 9 0 _ i 99
bo2-227-205- 222-238-248 766 1 27- 375 325 -14
k2 1 3 7 i>32-156-199 i,-04 2 5 11 i.-4 2 -21 I.-0 53 4 02 1.-53-204 K O 2 0 ff r>
k22 75 -90 t;33-190-16? 1.04 2 424 1-43-125 i.-4 24 80 u-054 392 1.-54-22E w 5 2 5 9
u-23 6 2 -32 w34-186-100 1.-04 4 44 3 i.-44 -50 i,-4 34 59 1.055 407 1.-55-192 w 5 4 5 1C
v24-129-158 1,-35-1 76 -7 1 1.-04 5 405 U-45-17G w4 4 4 34 i.0 5C 429 1.-5C- 1 58 w 5 5 5 18
w25 -39-110 i,-36-107 -65 i,-04 6 415 1.-4C-101 i.-4 5 4 50 i.-05 7 403 w 5 7 7 3 w o C cj 60
1.-26-298-246 u-3 7-202-15 5 w-04 7 379 W47-284 w4 6 4 18 1,-058- 419 i:58 -89 w5 7 5 -247
1.-27-210-150 1,-38-586-382 w04 8 401 w48-201 w4 74 27 w059- 172 w 5 S 4 1 7 w585 1042
1,28-245-274 u-3 9-281-281 w04 9 376 u-49-276 i,-4 84 19 1-0601140 i-*6 0101 w 5 9 C -910
v-29-1 24 -91 i,-40-169 125 v:050 421 w50- 134 V.-4 95 36 1.-0C1- 4 5 5 i.-ei o kGOC -209
w30 -165-14 9 u-4 1-240-131 w05 1 403 i,-51 -185 1.-505 29 u-062 85 i,-G2 -4 4 wG 16 -4 5
w31-216-171 w052 387 K52-241 i.-51 5 O
p to push/b to break/ to continue:
disp/ exam/ get/ save/ se t/ clear i cycle do log nei' rstart ] ptrain quit
reset run strain tail test
epoch 24 7 tss 3 7.691 ps s 1.3606 S = or 0 . 0000 cpname p.9 n 20
ta 20 2 2 20 50 7 9 9 7 97 79 50 50 20 rt o 20 50 79 97 ! 37 79 5 0
ah 41 49 35 25 16 1 1 0 3 4 14 22 30 25 19 8 10 48 0 0 1
ao 1 7 8 10 14 49 84 86 8 4 80 50 15 9 e 8 o 11 9 9 99 ; 15 50 52
do 0 0 0 0 0 0 1 1 0 0 4 0 0 0 1 2 0 0 9 7 0
bill -4 1 -10 -61 -59 -86 -21 -GC -23- 135 -SI - 31
bh2 -77 -4 7 - 7 - 118 -218- 176 491-316- 228- 233
bol- 260 -218 -256- 220 -229- 195- 221-200- 24 9- 232-199
bo2- 227 -205 -222- 238 -24 8 766 127-275 325 -14
i.-21 37 i.-32- 156 -199 1,-04 2 54 1 1.-4 2 -21 i,-05S 402 i:5 3 - 204 u*525 o o
w22 7 5 -90 w33- 190 -163 u-04 3 4 24 vJ3 -125 v4 2 4 80 i-Of.4 392 1.-5 4 - 228 v 5 3 5 c
i.-23 62 -32 w34 - 186 -100 1.-04 4 4 4 3 i.-4 4 -50 1.-4 3 4 59 w05 5 407 u-5 5 - 19 2 k54 5 16
w24 - 129 -153 w3 5- 176 -71 w04 5 4 05 i.-4 5 -176 k44 4 0 -1 w 1 u-056 429 w56- 158 V jS3 18
w25 -39 -110 w3G- 107 -65 1.-04G 415 i.-4 G -101 I.-4 54 50 i."057 403 i.-5 7 -73 1.-56 5 60
1.-26- 298 -246 i." 3 7 - 202 -15 5 u-04 7 379 i.-4 7 -2e4 1.-4 64 IS k0 58- 4 19 t.-58 -89 *c5 7 3 -247
1.-27- 210 -150 i.-38- 586 -382 1.-0 4 8 401 i.-4£ -201 i,-4 7 4 2 7 1.-059- 172 i,-59- 417 1.-585 1042
i.-28- 245 -274 w39- 231 -281 v 0 4 9 3 75 i.--19 -276 1.-4 84 19 1.-0601 140 i.-60- 101 1.-596 -910
i.-29- 12 4 -91 w4 0- 169 -12 5 I.-050 421 i.-50 -134 1,-4 9 5 36 1.061- A t o i.-61 0 w606 -209
u-30- 165 -149 i.-4 1 - 240 -13! 1.-051 4 03 i.-'l -185 -..*50 5 29 1.062 85 -..-6 2 -4 ! i.-G 1 6 _ J Tj
1.31- 216 -171 1.052 387 i.-o2 -241 i.-515 o


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 247 tss 39.253 pss 1.5G25 gcor 0.0000 cpname plO in 50
ta 50 20 2 2 20 50 79 97 97 79 50 50 20 2 2 20 50 79 97 97 79
ah 44 52 36 28 24 9 3 0 3 2 4 12 27 25 19 8 10 48 0 0 1
ao 50 15 9 10 14 50 83 85 84 79 48 14 9 8 9 11 99 99 45 50 52
do 00000001 1 0040000007106
bhl -41 -10 -61 -59 -8G -21 -66 -23-135 -91 -31
bh2--77 -47 -77-118-218-176 491-316-228-233
bol-260-218-256-220-239-195-221-200-249-232-199
bo2-227-205- -222-238-248 766 127- 375 325 -14
w21 37 w22-15G 199 w04 2 54 1 w4 2 -21 wO53 4 02 W53-204 tv a 2 a O o
w22 75 -90 i.'33-190-163 w04 3 124 w 13-125 w4 24 eo w054 392 v 5 4 2 2 S w535 Q
w23 62 -32 w3 4-166-100 w04 4 443 w44 -50 w4 3 4 59 tv05 5 4 07 tv5 5- 1 9 2 w 5 4 5 16
u-24-129-158 w35-176 -71 w04 5 405 w4 5 17 6 w4 4 4 34 w056 429 w5 6- 158 KDDO 18
w25 -39-110 w36-107 -65 w04 6 415 tv 4 6 1 01 w4 54 5 0 w057 403 tv 5 7 7 3 w5G5 60
w26-298-24 6 w37-202-155 w04 7 379 w 4 7 2 8 4 w4G4 18 W058-419 w58 -89 w5 7 5 -247
w27-210-150 w3S-586-382 w04 8 401 W48-201 w4 7 4 27 w059-172 w59-417 tv 5 8 5 1042
w28-245-274 w39-281-281 w04 9 376 w4 9 2 7 6 w4 84 19 w0601140 tv60-101 w59 6 -910
w29-124 -91 w4 0-169-125 w050 421 W50-134 w4 95 36 wOG1-4 55 w61 0 u-606 -209
w30-165-149 w41-240-131 w051 403 w51-185 w505 29 w062 85 w62 -44 w6 16 -45
w21-216-171 wO 5 2 387 w5 2-241 w515 3
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
ecoch 247 tss 41.092 pss 1 . 8381 gcor 0. 0000 cpr.a me p 11 in 79
ta 79 50 20 2 2 20 50 79 97 9 7 79 50 50 20 2 2 20 50 79 97 97
ah 47 50 41 27 26 25 11 1 3 2 2 4 15 23 19 8 10 48 0 0 1
ao 82 50 14 10 10 13 50 82 85 84 77 45 14 9 9 11 99 99 45 50 52
do 0000000 0 1 10 14 0 0 0 0 0 3 10 11
bhl -41 -10 -61 -59 -86 -21 -66 -23-135 -91 - 31
bh2 -77 -47 -77-118-218- 17G 491- 316-228-233
bol-260-218-256-220-239- 195- 221 - 200-249-232-199
bo2-227-205-222-238-248 766 127- 375 325 -14
w21 37 w32-156-199 tv04 2 541 tv42 -21 w05 3 402 tvE2-2C4 w 525 22
w22 75 -90 w33-190-163 w04 3 424 w 4 3 -1 2 5 tv 4 2 4 80 w054 392 t.-5 4-228 w 5 3 5 o
w23 62 -32 w34-186-100 w04 4 443 w44 -50 v4 34 59 tv055 407 I.-5 5-192 w o 4 o 15
w24-129-l58 W35-176 -71 w0 4 5 405 w45-176 w444 34 w0 5 6 429 w-56-158 w555 18
w25 -39-110 w3C-107 -65 w04 6 415 1.-46-10! w454 50 tv057 403 w57 -73 w 565- 60
w26-298-246 t.-37-202- 155 w04 7 379 1.-4 7-284 tv464 18 w058- 419 k 5 8 8 9 w 5 7 5 _ 2 i 7
w27-210-l50 w38-586-3e2 w04 8 401 w4 6-2 01 w4 7 4 27 w0 59- 172 tv 5 S 4 1 7 tv 585 1042
w28-245-27 4 w39-231-281 wO 4 9 27 6 w49-276 tv484 19 w0601140 t.-eo-ioi w 596 -910
W29-124 -91 w40-169-125 tv 0 5 0 421 t.-50-134 v4 9 5 36 w-OG 1 - 4 5 5 -..-51 0 w 606 -209
w30-165-149 w4 1 -240-131 wO 51 ' 403 wo 1-185 w505 2 9 tvO.62 3 5 w62 -44 tv 516 - 4 5
w31-216-171 w052 387 1.-52-241 w515 3


ptrain quit
p to push/b to brealt/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart
reset run strain tall test
epoch 247 * ss 4 o 913 pss 1 .8 214 gcor 0. 0000 cpname Pi 2 in
ta 97 79 50 20 2 o 20 50 79 97 97 79 50 50 20 O 0 20 50 79 97
ah 48 46 43 17 25 29 O *> 8 5 2 1 o 5 13 17 8 10 48 0 0 1
ao 92 64 51 15 10 9 13 50 81 84 62 76 43 14 10 11 99 99 45 50 52
do 0 0 0 0 0 0 0 0 0 1 n 0 1 4 0 0 0 0 -1 5 11
bhl -4 1 -10 - c: -59 - 86 - 21 -6G - 23- 135 -91 - 31
bh2 - 7 7 -47 - 7 7 -118 _ o 18-1 7 6 491 -316- 228- 233
bo 1-260-218-256-220-239-195-221-200-249-232-199
bo2-227-205' -222-226-248 766 127- 375 325 -14
w21 3 7 i. 32-156-199 w04 2 54 1 i-42 -21 w0 52 402 1.-5 3-204 w 5 2 5 O 0
w22 75 -90 w33- 190-163 1.-04 3 424 w4 j-12a w4 2 4 80 1.-054 29 2 1-54-228 w5 2 5 9
w23 62 -32 1.-34-186-100 1.-04 4 443 1.-4 4 -50 i.-4 34 59 i.*055 4 07 i-55 l 92 Will 0 16
w-24-129-158 1.-35-176 -71 w04 5 405 w-1 5-176 w4 4 4 34 1-056 4 29 1.-5C- 158 0 %) D 18
w25 -39-110 V3G-107 -65 w04 6 415 w-46-101 w4 54 50 1-057 403 i.-57 -7 3 w f> G 5 GO
w26-298-246 w-3 7-202-155 u-0-17 379 1.-47-284 w4 6 4 ie 1.-05S-419 k 5 8 8 9 w 5 7 5 -247
i.'27-210-150 1.38-58G-3 82 w-04 8 "4 0 1 1.-46-201 1.-4 7 4 27 1.-059-172 1-59-417 v5 8 5 1042
w28-245-274 w39-281 2.81 1.-04 9 376 w49-276 w-4 84 19 1.-0601140 u-CO-101 v5 9 6 -910
w29-124 -91 w4 0-169-125 w050 421 w50-134 1.-4 95 36 i.-06 1-4 55 i.-61 0 w606 -209
w30-165-14 9 k4 1-24 0-131 w051 403 w-5 1 -185 w505 29 wOG2 85 i.-62 -4 4 w6 ] 6 -4 5
w-21-216-171 w052 387 w-52-241 w515 3
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain
reset run s train tall test
epoch 247 tss 44.434 pss 1 .5213 gcor 0. 0000 cpname pi 2 in 9 7
ta 97 97 79 50 20 2 2 20 50 79 97 97 79 50 50 20 2 2 20 50 79
ah 48 43 42 6 22 26 25 26 10 3 1 i 2 a 11 8 10 48 0 0 1
ao 92 91 85 52 15 9 9 12 43 80 83 62 76 40 13 11 99 99 45 50 52
do 0 0 0 0 0 0 0 0 0 0 2 2 0 2 4 0 0 0-5 0 6
bhl -41 -10 - 61 -59 -86 - 21 -CG -23- 135 -91 - 31
bh2 -77 -47 - 77-118-218-1 76 491 -316- 226-233
bo1-260-218-256-220-239-195-221-200-249-232-199
bo2-227-205- -222-238-248 766 127- 375 325 -14
w21 37 -32-156-199 v04 2 541 v-4 2 -21 w053 402 '.-53-204 wa2 5 22
w-22 75 -90 v33-190-163 w04 3 424 '.-4 3-125 w4 24 80 w054 392 w54-228 wa 3 5 9
w23 62 -32 t." 3-i 18c 100 w04 4 443 w44 -50 w4 34 59 v055 407 i.-5 5-192 w 5 4 5 16
V24-12S-158 w35-176 -71 w04 a 405 w 4 a 1 i 6 w4 44 3 4 wOaa 429 w a 6 1 5 8 ^300 18
w25 -39-110 w36-107 -65 w04 6 415 w46-101 w 4 54 50 w057 403 w57 -73 w 5 6 5 60
..26-298-246 '37-202-155 w04 7 379 w47-284 w4 64 18 '.058-4 19 waB -89 '575 -247
..27-210-150 '.-38-586-382 w048 401 u-48-201 w4 7 4 27 '.059-172 wa9-4 17 '.585 1042
u-28-245-274 '.-39-281-281 '.04 9 3 76 ..4 9-2 76 w4 84 19 wOSOl140 '.60-101 wa96 -910
..-29-124 -91 '.40-169-125 w050 421 koO-134 '.*4 9 a 36 w061-455 wo 1 0 w606 -209
w30-155-14 9 '4 1-240-1 31 .-051 403 '-51-185 w5 05 29 w062 55 '.-62 -4 4 '.*c: 6 -i w
'31-216-171 u-052 387 '.-52-24 1 '.515 3


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 247 tss 45.581 pss 1.1467 gcor 0.0000 cpname P14 in
ta 79 97 97 79 50 20 2 2 20 50 79 97 '97 79 50 50 20 o O 20 50
ah 47 39 42 3 13 9 23 30 15 11 4 1 2 3 5 6 9 47 0 0 i x
ao 82 91 90 84 53 18 10 8 12 45 80 83 82 75 37 15 99 99 45 50 52
do 0 0 0 0 0 0 0 0 0 10 2 2 0 3 4 0 0 -6 - 7 0
bhl -41 -10 - Cl -59 -86 - 21 -66 -23-135 -91 -31
bh2 -7 7 -47 - t 1 -118-218-1 7 C 491-316-228- 233
bol-260-218-256-220-239-195-221-200-249-232-199
bo2-227-205- 222-238-248 766 127- 375 325 -14
w21 3 7 <-32-156-199 <-04 2 541 <-4 2 -21 <-053 402 <-5 3-204 w5 2 5 O
w22 75 -90 <33-190-163 w04 3 4 24 <-4 3-125 <-4 2 4 80 w054 392 <-54-228 v53 5 9
i-23 62 -32 <-34-186-1 00 <-'04 4 4 4 3 <-4 4 -50 <-4 3 4 59 <-055 407 w 5 5 -1 9 2 w54 5 1G
w24-129-158 w35-176 -71 w04 5 405 . <-'4 5-176 w4 4 4 34 w05C 429 <- 5 6-158 W 5 1 o
w25 -39-1 10 <-36-107 -65 <-04 C 415 <-46-101 <-4 54 50 w057 403 w57 7 3 w56 5 60
V20-293-24C <-37-202-155 <-04 7 379 <-47-284 <-4 64 18 <-058- 4 19 <-58 -89 w 5 7 5 -24 7
w27-210-150 <-38-586-382 <-048 401 <-48-201 <-4 74 27 <-059- 172 w59-417 w585 1042
w28-24 5-27 4 w3 9-281-281 <-04 9 376 <-49-276 <-4 84 19 <-0601 140 <-60-101 <-5 96 -910
w29-124 -91 <-40-169-1 25 <-050 421 <-50-134 w4 95 36 <-061- 4 55 <-61 0 w60 6 -209
v30-165-149 w4 1-240-131 <-051 403 <-51-185 w505 29 <-062 S 5 w62 -44 <-6 1C -45
<-31-216-171 <-05 2 387 <-52-24 1 <-515 3
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ] ptrain qu
reset run strain tail test
epoch 247 tss 46.398 pss 0.816-7 gcor 0. 0000 cpn ame pl 5 i: n 5 0
ta 50 79 97 97 75 50 20 2 2 20 50 79 97 97 79 50-50 20 2 2 20
ah 44 36 40 2 9 1 10 26 17 20 16 3 2 3 3 3 5 4 1 0 0 1
ao 50 84 90 88 84 56 18 9 9 12 43 79 82 82 7 5 38 99 99 45 50 52
do 0 0 0 0 0-10 0 0 0 10 2 2 0 2 0 0 -4-11 -7
bhl -41 -10 -61 -55 -86 _ rt - -66 -23-135 -91 - 31
bh2 -77 -47 -77-118-218- 176 491-316-228- o n o
bo 1- 260-218- 256-220-239- 195- 221-200-249- O 0 n 1 99
bo2-227-205- 222-238-248 766 127-375 325 -14
w21 3 7 <-32-156-199 <-04 2 541 <--4 2 -21 <05 3 402 w53-204 <-5 2 5 2 2
w22 75 -90 W33-190-163 w04 3 4 24 w-. 8 125 w4 2 4 90 w0 54 392 <-5 4-228 <-5 2 5 Q
<-23 62 -32 <-34-186-100 <-04 4 443 w4 4 -50 <-4 3 4 59 <-055 407 <-55-192 <-5 4 5 16
w24-129-158 w35-176 -71 <-04 5 405 w45-176 w4 4 4 34 w05C 429 w56-158 w555 1 o
<--25 -39-110 w36-107 -65 wO 4 6 4 15 w4 6-101 w4 54 50 w057 403 w 5 7 7 3 w5C5 60
w2 6-298-246 <-37-202-155 <-04 7 279 w4 7-284 w 4 6 4 18 <-058 -419 <-58 -89 w5 7 5 -247
-27-210-150 <-38-5 36- 382 <-04 8 401 <-48-201 <-4 7 4 2 f <059 -172 <59-417 <-585 1042
v28-245-274 <-39-281-281 w04 9 376 w! 9-275 w4 84 1 9 wO'60 114 0 w60-101 <-596 -910
<-29-124 -91 <-40-1 69-125 -050 421 <--50-134 <-4 9 5 36 <-061 i w <-£ 1 0 <-60 6 -209
<30-16 5-149 <-4 1-240-1 3 1 <:05 0 3 <--51-185 <-505 2 9 <-05 2 O <-C2 -44 <:6 1 6 1 1 w
-31-213-171 <-052 337 <-*52-241 w515 3


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset rur: strain tall test
epoch 247 tss 47.362 pss 0 .9648 gcor 0.0000 cpname plG 4
ta 20 50 79 97 97 79 50 20 2 2 20 50 79 97 97 79 50 50 20 2 9
ah 41 38 36 3 8 0 3 9 15 2 9 10 5 3 3 o 1 15 0 0 1
ao 17 49 85 89 88 83 57 19 9 9 11 41 79 83 83 77 98 99 4 5 50 52
do 0 0 0 0 0 0 -1 0 0 0 0 2 0 2 2 0 0 0 0- 10- 1 0
bhl -41 -10 -G1 -59 -86 - 21 -66 -23- 135 -91 -31
bh2 -77 -47 -77-118-218-17G 491 -31G-228-233
bo 1 -260-218-25G-220-239-195-221-200-2 49-232-199
bo2-227-205- -222-238-248 7G6 127- 375 325 -14
w21 3 7 w 3 2 15 6 19 9 w04 2 54 1 w-12 -21 '.053 402 w53- 204 w i, 2 ij '! O
w22 75 -90 w33-190-163 wO-13 12 1 w43-125 w 124 80 '.054 392 '.54-228 k5 3 5 9
i-33 62 -32 u-34- 186-100 w04 4 443 w4 4 -50 '4 3 4 59 w055 407 w55-192 w54 5 16
'.'24-129-158 v3 5-1 76 -71 w04 5 405 '4 5-176 '.4 4 4 34 i.056 4 29 w56-158 k5 5 5 18
w25 -39-110 w3G -107 -65 w04 6 415 W4G-101 w4 54 50 w057 403 w 5 7 7 3 v5 G ri GO
w26-298-246 w 37-202-155 w04 7 379 >.47-284 w4 64 18 wO58-4 1 9 w5B -89 w f> 7 5 -247
V.-27-210-150 W38-58G-382 w04 8 401 w48-201 w4 7 4 27 '.059-172 w59-417 w58 5 1042
w28-245-274 v39-281-281 w04 9 376 w4 9-27 6 w4 84 19 w0601140 w60-101 w59G -910
'.29-124 -91 w4 0-16 9 125 w050 421 wSO-134 w4 95 36 i.OG 1 4 5 5 wGl 0 wBOG -209
w30-l6 5-14 9 w41-240-131 wOf 1 403 '.51-185 '.505 29 w062 85 wG2 -44 w6 1 6 -45
w31-216-171 w052 387 w52-241 w515 3
p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 247 tss 48.623 pss 1.2600 gcor 0.0000 cpname pl7 in *2
ta 2 20 50 ' 79 97 97 7 9 50 20 2 2 20 50 79 97 97 79 50 50 20 2
ah 40 42 33 5 8 0 1 1 9 19 32 21 16 6 3 2 0 1 0 0 1
ao 7 15 49 85 88 86 82 58 17 10 8 11 38 78 84 84 89 99 4 5 50 52
do 0 0 0 0 0 1 0 -2 0 0 0 0 2 0 1 2 0 0 4 -6 _ 1 o
bh : -41 -10 -61 -59 -86 -21 -66 -23- 135 -91 - 31
bh2 - 7 7 -4 7 -77- -118 -218 -176 491- 316- 226- 2 3.
bo 1 - 260 - 21 8 - 256- -220 -239 -195- 221- 200- 249- 232-199
bo 2- 227 -205- 222- -238 -248 766 127- 37 5 325 -14
w21 37 w32- -156 -199 w04 2 541 '.42 -21 w05 3 402 Zi 3 204 v5 25 22
w22 7 5 -90 w33- -190 -163 '.04 3 424 v-4 3 -125 '.4 24 80 '.054 392 w54- '.-535 9
w23 61 -32 w34- -186 -100 '.04 4 443 w4 4 - 50 v4 34 59 w05 5 407 w5 5 - 192 v54 5 16
w24- 129 -158 w3 5- -176 -71 w04 5 405 w4 5 -176 w44 4 31 w056 4 29 w56- 158 w5 5 5 18
'.-25 -39 -110 w36- -107 - 6 5 w04G 415 w4 6 -101 '.-4 54 50 w05 7 403 '.*o 7 -73 w 5 5 5 60
w26- 298 -246 w3 7- -202 -15 5 '.04 7 379 i-4 7 -284 '164 18 '.058- -419 '.-58 -89 7 5 -247
'.2 7- 210 -150 -38- -586 -se2 '04 e 401 w4 6 -201 >.4 7 4 27 '.059- -172 w59- 417 k 5 S 5 1042
w28- 245 -274 >.-39- -281 -281 '.04 9 376 w4 9 -276 w484 19 w06C: 1140 w60- 101 w 5 9 G -910
'.29- 124 -91 w4 0- -169 -125 >.050 421 w 5 0 -134 w4 95 36 wOGl- -4 55 w61 0 u-60G -209
--3 0- 165 -149 v4 1- -24 0 -13 2 t.Coi 103 '.5 1 -185 '-505 29 '.0C2 u D '.62 J J v6 : G
'.21- 216 -171 '.052 387 k 5 2 -24 1 w 515 *_>


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain cuit
reset run strain tall test
epoch 24 7 tss 48.909 pss 0.28G1 gcor 0.0000 cpriame pi 8 in
ta 2 2 20 50 7 9 9i 9 i 79 50 20 2 n 20 50 79 9 7 97 79 50 50 20
ah 40 45 34 1'4 10 0 1 0 4 9 28 23 27 17 G 2 0 1 0 0 1
QO 7 9 14 49 84 87 8G 81 57 17 9 9 11 35 79 85 84 99 45 50 52
do 0 0 0 0 0 1 1 0 -1 0 0 0 0 3 0 1 1 0 0 0 - 7
bhl -41 -10 - G1 -59 -8G -21 -GG -23- 135 -91 -31
bh2 -77 - A i - 7 7- 118-218- 1 7 G 491-31G- 228- 233
bo 1-260 -218-2 56- 220-239- 195- 221-200- 249- 232 -199
bo2-227-205- -222-238-248 7GG 127- 375 325 -14
w21 37 w32-15G-199 w0 4 2 541 w42 -21 w053 402 1.-53-204 U M fi r%
w22 75 -90 w3 3-!90-16 3 w04 3 4 24 w4 3 -12 5 wl 24 80 w054 392 w54-228 i.* 5 3 5 <,
w23 G2 -32 W34-186-100 w04 4 443 w44 -50 -4 34 59 w055 407 w5 5-192 K 5 4 5 10
w24-129-158 w35 17G -71 w04 5 4 05 wl 5-176 w 4 4 4 34 w056 429 w5G-l53 k 5 5 5 13
w25 -39-110 w36-107 -G5 w046 415 w 4 G -101 w4 54 50 w05 7 403 w a i / 3 w 5 C 5 GO
w26-298-246 w3 7-202 -15 5 w04 7 379 w4 7 284 w4 6 4 18 w058- 419 w58 -89 u 5 7 5 -24 7
w27-210-150 w38-58G-382 w04 8 401 w4 8-201 w4 7 4 27 w059- 172 w59-417 w585 1042
w28-24 5-274 w39-281-281 w04 9 37 G w 19 2 7 6 w4 84 19 wOGOl140 wGO-101 k59G -910
w29-124 -91 w40-169-125 -050 421 w50-134 w4 95 36 wOG 1 - 4 5 5 wC 1 0 v.*60G -209
w30-165-149 w4 1 24 0-131 w051 403 w51 -185 w5 05 29 wOG 2 8 5 wC2 -44 wGIG - 4 5
W31-21G-171 wO 5 2 387 w52-241 w515 3
p to push/b to break/ to continue:
disp/ exam/ get/' save/ s< et/ clear i cycle do log newstart ptrair i qu
reset run strain tall test
epoch 247 tss 49.7G7 ps s O.S'588 SCOT 0. 0000 cpname p 19 in 20
ta 20 2 2 20 50 79 97 97 79 50 20 2 2 20 50 79 9'7 97 79 50 50
ah 41 49 35 25 16 1 1 0 3 3 12 21 29 24 14 3 0 1 0 0 1
ac 17 8 10 14 49 84 86 85 81 56 17 § 9 11 33 78 85 99 45 50 52
dc 00000011 0-10 0 0 0 3 0 1 0 8 0 0
bhl -41 -10 -61 -59 -86 -21 -66 -23- 135 -91 - 31
bh2 -77 -47 -77-118-218-176 491-316- 228- 233
roi-260-218-256-220-239-195- 221-200- 249- 232-199
fcc2-227-205-222-238-248 766 127-375 325 -14
w21 27 w22-l56-199 w042 541 v4 2 -21 w053 402 w53-204 w525 2
w22 75 -90 W33-190-1G3 w043 4 24 w43 -125 w 4 2 4 80 v054 392 V54-223 w533 o
-23 62 -32 w34-18G 100 ---044 443 w44 -50 w 4 3 4 59 w05 5 407 w55-122 w545 16
w24-129-158 w35-17C -71 w045 405 w45 -176 w-i 4 4 34 wO 5 6 429 w 5 G -15 8 w 5 5 5 18
w25 -39-110 w36-107 -65 w046 435 w46 -101 w4 5 4 50 w05T 403 wa i i 3 wo6 a 60
v26-298-246 w37-202-155 w047 379 w47 -234 v4 64 18 w058- 419 w58 -89 w575 -247
w27-210-150 W38-5SC-382 w04S 4 C1 w4 8 -201 w-i 7 4 27 w059- 172 w 5 9 4 1 7 w 5 8 5 1042
w28-245-274 v39-281-281 w049 376 w4 9 -276 w4 84 19 w0601 140 w60-101 w59C -910
..29- 124 -91 w4 0 -1 6S- 1 25 w050 421 w50 -134 w4 9 5 36 wOG 1 - 4 0 0 wG1 0 wCOC -209
w 2 0 : 6 5 1 ! 9 w4 1-2 10-1 31 w05j 103 w51 -185 w505 29 w062 85 i.-62 -4 4 ' i 1 'J -4 5
w 31-216-171 I.-D5 2 33 7 i.-52 -241 w5 1 5 3


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 24 7 tss 50.87 3 ps s 1.1054 gcor 0. 0000 cpiiame p20 in 50
ta 50.20 2 2 20 50 7 9 97 97 79 50 20 2 2 20 50 7 9 9 7 97 79 50
ah 44 52 36 28 24 9 3 0 3 2 3 1 1 26 25 18 G 1 2 0 0 1
ao 50 15 9 10 14 50 83 85 84 80- 54 16 9 9 12 32 95 99 45 50 52
do 0 0 0 0 0. 0 O' 1 10-10 0 0 1 4 0 0 9 7 0
bill -41 -10 -G1 -59 -86 -21 -6C -23-135 -91 - 31
bh2 -77 -47 -77-118-218 -176 491-316-228- 233
bol-260-218 -25C-220-239 -155- 221-200-249- 232-199
bo2-227-205 -222-238-248 7 C 6 127-375 325 -14
w21 3 7 1.-32-156-199 wO 4 2 541 w42 -21 w05 3 402 w53-204 w5 2 5 n 'i
w22 75 -90 v 3 3 1 9 0 1 6 3- w04 3 424 W43-125 w 124 80 vG54 392 1.-54-228 i-535 9
w23 62 -32 w 34-186-100 w04 4 443 w44 -50 v4 3 4 59 w 0 5 5 407 w55-192 w 5 4 5 16
w24-l29-158 w35- 1 76 -71 w04 5 405 w45-176 w4 4 4 34 w 0 5 6 429 w 5 6 1 5 8 1-555 18
w25 -39-110 w3G-107 -65 w04G 415 w46-101 w 4 5 4 50 w05 7 403 i-5 7 7 3 1-565 60
W26-298-24G w 37-202-155 w04 7 379 w47-284 w4 64 18 w058- 419 w58 -89 i-5 7 5 -247
k2 7-210-150 w38-586-362 w04 8 401 w4 8-201 w4 7 4 27 u-059- 172 u-59-41 7 v585 1042
w28-245-274 w39-281-281 w04 9 376 w49-276 w4 84 19 w0601 140 1.-60-101 i.-5 96 -910
w29-124 -91 w4 0-169-125 w050 421 wSO-134 w495 36 wOG 1 - 455 i-61 0 wGOG -209
w30-16 5 -14 9 w41-240-131 w05 1 403 w51-185 w50 5 29 wOG 2 85 1.-62 -4 4 i-6 1G -4 5
w31-216-171 w052 387 w52-241 w515 3


Impulse.str file for sinusoidal response with desire outputs c-f
delay network as inputs.
get network impulse.net
get patterns impulse.tpt
set nepochs 100
set ecrit .1
set dievel 3
set slevel 1
set .1 flag 1
set mode follow I
set mode lgrain epoch
set param irate .5
get weight impulse.wt s
n-sample


^ *1
lmpulse.net file
definitions
nun it s 2 3
ninputs 11
noutputs 1
end
network:
Tir 11 1 0 2
% 12 1 0 12
% 13 1 1 12
rrr. % H 1 2 12
% 15 1 2 I 2

ic 1 4 12
rrr. .
% 17 1 5 12
rrr. % 18 1 6 * */
% 19 1 7 i 2
71 20 1 e 12
/H M 1 : 9 12
Sr 22 1 11 11
end
biases:
%r 11 12
end


mpulse.tem file
define: : layout ) 7 131
epoch input s tss S pss S C gcor S cpnaae £
actout acthid delhid biahid s tur£ s 5 S delout S biacut £
i.-U e
wl2 S
wl3 c*
wn c
Wl 5 s
wlG s
wl 7 c
wlB c
w! 9 c
w20 c-
i.-21 S
w22 end e
epochno variable i s n epochno 6 1.0
tss floatvar i n tss G 1.0
SCOT f loatvar 2 S 3 gcor G .1.0
cpname variable o s n cpname -4 1.0
pss iloatvar o c 2 pss 6 1.0
input ve:ior 2 c 5 act:vati on h G 1000 .0 0 1 1
acthid vector o s 10 activation h 5 1000 .0 11 1 1
actout vector 2 s G act i vation h 5 1000 () 1
target vector o s n target h 5 1000 .0 0 1
delhid vector r> s i: delta h 5 1000 .0 11 1 i
delout vector f> g £ delta h g 1000 .0 22 2
biahid vector O c 13 bias h 5 100. 0 11 i 2
biaout vector o fr o bias h 0 100. 0 22 1
wll matri:: 3 13 weight h 4 100. 0 11 1 0 2
wl 2 matri:: 3 s n weight h 4 100. 0 12 i 0 " 2
13 matri:; 3 s r; weight h 4 100. 0 13 2 i 12
wl4 matri:: 0 o n weight h 4 100. 0 14 i o 1 o x _
wl 5 matrix 0 w* e n weight k 4 : 00 . 0 IS 1 o sJ 1 o
wl 6 ma t r r 2 c n weight h 4 100. 0 15 i 4 r.
w 1 7 matrix 3 c* n weight h 100. 0 17 i 5 12
w 1 8 matri:: 3 S n weight h 4 100. 0 IS 6 12
wl9 matrix 3 g n weight h 4 100. 0 IS I 7 12
w20 matrix O s n weight h 'i 100 . 0 20 1 o 12
2 matrix* 0 c ii weight h 4 100. 0 21 1 o 12
w22 matrix 2 £ n -eight h 1 100 . 0 22 1 1 i 2 1


Impulse.spt file with desire sinusoidal outputs of n-sample delay network
11 co o n -y o> 1- in o
co T o o o O m "7 Cl in
O O o o -T i- oi C) i- o co m m Cl o o a*
o Cl i- i- C) o o Cl Cl o
o o o o o o o o o o m 1- C C) 1- in CM o o Cl
o o o o O o o o O o o o
o c. m m CD o T *V o
o Ol i- i- Cl o o Cl CM o o
o o o c o o o o m 1- CD CD 1- lO CJ o o Cl in
o o o O O o o o o o o o
o o o o o o o C: o CD IvD in CO o SI* o n
o Cl 1- i- Cl o G Cl Cl c o C)
m 1- 1 CD in Cl o o Cl m 1-
o o o o O o o o o o o o o o O o o o o o
o co in m CO o o 3* CO o 7; m
o i- i- CD o o Cl Cl o o Cl i
in i - OI oi 1- in CJ o o Cl m 1- oi
o o c o o o o o o o a o o o o o o o o CJ
o CO m m CJ o o to o Cl 1.1 in
o CD i - i- C: o o CM CM o o Cl 1 - i-
in OI CD i m Cl o o Cl l.'i 1- Oi Cl
o o o o o o o o o O O o o o => o o o o o
o CO m m CD o o o o c> i- i- CD o o CM Cl o o Cl i- 1 - o>
in I- Cl Oi 1- m Cl o o Cl in 1" 01 O 1-
o o o o o o o - o O o o o o 3 o G o o o
o CO m m co O o 77 -T o o CD in m CO o
o CD i- i - OI o o CM Cl o o Cl 1 - i- O) o
m 1- O) c> 1- m Cl o Cl 1- m
o o o o o o o o O o O o o o o o o o o
o co m m CO o o -T o CO m in C) a
o od i-- i- Cl o o Cl CM o o CD i- i- Cl o o
m i- OI C) I** m Cl o O Cl m I- Oi OD 1- in Cl
o o o o o o o o o o o O o o o o O o o o
o CO m in co o T O o CO in m CD o 51*
o Cl i - i- Cl o o CM Cl O o Cl I-- i Cl o o CJ
M 1- oi o i- m Cl O o C 1 m 1- OD Cl 1 i:i Cl o
o o o o o o o o o O o G o o O o o o 3 o
o co in m CO o -T to O CO m in CO o T T
o Cl i~ r- CD o o Cl Cl o O CD i- i- Cl o o CM CJ
m 1- o Cl I'- m Cl o o CJ in 1 - CTl Cl 1"- m Cl O o
o o o o O o o o o o o o O o o o o o 3 o
o CO m in to o o T T o co co in OI in OD co CO o OI co C) -r l- -l* 1- to
o 1- i- OI O O OI CJ o o c: CD i- i- O Cl ci o Cl o CM Cl CM Cl o
If) 1 O) o o o O o o o o o o o o O o o o o O o o o o o o O o o o o
O f cj co t in - oq o o rH Cl CJ rr in 1- r 1 rt r < r | r< -I l | i CJ
aC'OiP.ap.aau.e'p' a P. A p. P, P, P, fj. p. a
500 C.20G C.C24 0.02-5 0.205 0.500 0.703 0.975 0.975 0.793 0.500


Initial impulse response network displays with desire sinusoidal outputs
of n-sample delay network as inputs.
p to push/b to break/ to cont inue :
disp/ exam/ get/ save/ set/ clear cycle do log newstart
reset run strain tall te St
epoch 0 tss 0.4353 pss 0.4353 gcor 0.0000 cpname
input 500 0 0 0 0 0 0 0 0 0 0
actout G90 target 31 delout -140 biaout
acthid 378 G 1 3 358 -171 484 411 374 511 525 411 507
delh id -G -5 -13 2 1G 3 7 7 0 0 -8
biahid -33 If. -38 -14 - : 4 -37 -34 5 33 _ n .i 10
wll -33 1
w 1 2 7 - 3 2 -3 7 0 0 0 0 0 0 0 c |
w] 3 -19 -3G -20 0 0 0 0 0 0 (I 0 -31
w 1 4 -4 5 -1 2 7 0 0 0 0 c 0 0 0 9
wl 5 4 3 -19 -30 0 0 0 0 0 0 0 0 18
w 16 25 -21 41 0 0 0 0 0 0 0 0 4
wl 7 -34 -38 _ i 0 0 0 0 0 0 0 0 -40
w 18 -42 -20 41 0 0 0 0 0 0 0 0 -1
wl 9 -28 45 16 0 0 0 0 0 0 0 0 -45
w20 1 7 4 5 -34 0 0 0 0 0 0 0 0 .09
w21 18 37 0 0 0 0 0 0 0 0 0 -17
w22 20 18 43 -3 -4 G -8 22 2 2 2 0 23
p to push/b to break/ter> to continue:
disp/ exam/ get/ save/ set/ clear cyci e do lo s new s ta rt
reset run strain tail test
epoch 0 tss 0.8447 pss 0.4094 gcor 0.0000 cc name
input 793 500 0 0 0 0 0 0 0 0 0
actout 688 target 4 9 delout -1 3 7 biaout
acthid 405 579 338 47 1 484 411 3 74 51 1 525 411 507
aelhid -e -5 -i3 2 16 3 -7 7 C 0 _ o
biahid -32 45 -38 -1 4 -14 -37 34 5 33 - ^ *1 10
wll -33 4 2
.. O 7 -32 -27 0 0 0 0 0 0 0 0 - 7
wl 3 -19 -3G -20 0 0 0 0 0 0 0 0 -31
1 4 -45 -1 27 0 0 0 0 0 0 0 0 9
w 15 43 -19 -30 0 0 0 0 0 0 0 0 18
wl 6 25-21 41 0 0 0 0 0 0 0 0 4
wl 7 -34 -38 -1 0 0 0 0 0 0 0 o -40
w 13 -42 -20 41 0 0 0 0 0 0 0 0 _ i
wl9 -28 45 18 0 0 0 0 0 0 0 0 -4 5
i.-20 17 45 -34 0 0 0 0 0 0 0 0 _ r> 1
w21 18 37 0 0 0 0 0 0 0 0 0 -17
w2 2 20 15 42 -2 -j r o 2 2 r> o o r\ 2 2
ptrain quit
p.O
! 1
ptrain quit
11


p to push/b to break/ to continue:
disp/ exam/ set/ save/ set/ clear cycle do log newstart ptrain qui
reset run strain tall test
epoch 0 . tss 1. 3006 pss 0 .4 559 gcor 0.0000 cpname P-2
input 975 7 93 500 0 0 0 0 0 0 0 0-
actout C83 target 8 del out -1 4 G bi aout 41
acthid 4 21 5 1 2 292 414 481 4 11 3 7 4 511 5 2 5 411 5 07
delhid 7 -5 -12 O 17 3 7 7 0 0 -8
biahid -32 4 5 -36 -14 -14 - 3 7 -3 4 5 3 3 -- 4 10
.-ll 33 -12
wl2 7 -32 -37 0 0 0 0 0 0 0 0 - 7
i.-13 19 -36 -20 0 0 0 0 0 0 0 0 -31
w 1 4 15 -1 27 0 0 0 0 0 0 0 0 n
wl 5 43 -19 -30 0 0 0 0 0 0 0 0 16
u-16 r 21 41 0 0 0 0 0 0 0 o
i-lT 34 -38 -1 0 0 0 0 0 0 c 0 -40
k18 4 2 -20 41 0 0 0 0 0 0 0 0 -1
wl9 28 45 18 0 0 0 0 0 0 0 0 -4 5
1.-20 17 45 -34 0 0 0 0 0 0 0 0 _ n o
i.-21 IS 27 0 0 0 0 0 0 0 0 0 -17
i.-22 20 18 43 -3 -4 6 -8 22 o 0 23
p to push/b to break/ t o con tinue:
disp/ exam/ get/ save/ set/ clear cycle do 1 os new start ptra
reset run su* ain tali t est
epoch 0 tss 1.7535 PS s 0.4529 geo r 0.0000 pname p.3
input 9 7 5 S 75 793 500 0 o o' 0 0 0 0
actout 67 o 4. target 0 delout -148 biaout 41
acthid 440 4 7C 246 379 534 412 374 511 525 4 11 507
delhid - 7 -5 -11 2 17 3 -7 - 8 0 0 -8
biahid - 3 2 45 -38 -14 -14 -37 -34 0 33 r. 1 10
wl 1 3 3 42
u-12 7 -32 -37 0 0 0 0 0 0 0 0 - 7
-13 2 9 -36 -20 0 0 0 0 0 0 0 0 -31
wl 4 4 5 -1 2 7 0 0 0 0 0 0 0 0 9
wl 5 n w -19 -30 0 0 0 0 0 0 0 0 18
vl6 o r . o' -21 41 0 0 0 0 0 0 c 0 1 T
i-17 34 -38 -1 0 0 COO 0 0 0 -40
wl 3 42 -20 41 0 0 0 0 0 0 O' 0
wl 9 23 45 18 0 0 0 0 0 0 0 0 -45
w20 17 4 5 -34 0 0 0 0 0 0 0 0 _oo
w 1 i 8 0 ( 0 C : 0 0 0 0 0 0 -1 7
w2 2 20 *. Cl 4 3 3 - 46 o r* o n o o 0 r* "2
qui


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain qui
reset run strain tall test
epoch C tss 2.0718 pss 0.2 183 geo r 0.0000 cpname
i nput 7 93 975 975 793 500 0 0 0 0 0 0
actout 6G9 target 105 delout -124 biaout
ncthid 4 55 449 224 290 542 443 371 511 525 411 507
delhid -G -4 -9 1 14 3 -G G 0 0 -7
biahid - 32 45 -38 -14 -14 -37 -2:4 5 33 -24 10
wll 33 42
w 1 £ 7 -32 -37 0 0 0 0 0 0 0 0 -7
u-1 2 ] 9 - 3 G -20 0 0 (' 0 0 0 0 0 -31
w 1 4 4 5 -1 27 0 0 0 0 0 C 0 0 9
wl 5 43 -2 9 -30 0 0 0 0 0 0 0 0 18
w 1 6 25 -21 41 0 0 0 0 0 0 0 0 4
K 1 7 34 -38-1 0 0 0 0 0 0 0 0 -40
k18 4 1. -20 41 0 0 0 0 0 0 0 0 -1
w 1 & 28 4518 0 0 0 0 0 0 c 0 -45
k20 17 45-34 0 0 0 0 0 0 0 0 -22
k21 18 37 0 0 0 0 0 0 0 0 0 -17
w22 20 18 43 -3 -4G -8 O O 0 o 2 0
p to push/b to break/ to continue
disp/ exam/ get/ save/ set/ clea r cycle do ii og newstart
reset run strain tail test
epoch 0 tss 2.1432 pss 0.0 713 geo r 0.0000 crjnair.c
input 500 723 S 7 3 975 793 500 C 0 0 0 0
actout 6 71 target 404 delout -58 biaout
ac-thid 1 -t 61 459 223 408 510 435 333 511 525 -ill 507
delhid -2 -2-4 0 6 1 _ 2 3 0 0 -3
biahid - 32 45 3c -14 -14 -37 -34 5 33 -24 10
wll o o 42
wl2 7 -32-37 C 0 0 0 0 0 0 0 -7
wl3 :o -3G -20 000 o r 0 0 0 -31
wl4 *15 27 0 0 0 0 c 0 0 0 9
wl 5 4 3 -15 -30 0 0 0 0 0 0 0 0 16
v-lC 25 -21 41 0 0 0 0 0 0 0 0 4
W 1 7 34 -38 -1000 0 0 9 0 a in U *i u
u-18 4 2 -20 41 0 0 0 0 0 C 0 0 -1
v-19 23 45 IS 0 0 0 0 0 0 0 0 -45
v2 0 17 45-24 0 0 0 0 0 0 0 0 22
u-21 1 c 27 0 0 0 0 0 0 0 f. 0 -17
v:2£ 20 18 42 -2 -46 -8 f. O Of; 2 r-
-11


p to push/b to break/ to continue:
ci sp/ exam/ set/ save/ set/ clear cycle do log newstart ptrain
reset run strain tall test
epoch 0 t£S 2.1500 pss 0.0069 geer 0.0000 cpname p.6
input 206 500 792 975 975 7 9 3 5 00 C i 0 0 0
actout 6 70 target 75 3 delout 18 b iaout 41
actnid J 5-1 49-1 243 441 480 481 267 45S i 531 4 10 507
delnid 0 0 10 -2 0 0 C i 0 0 1
biahid - 32 4*. -38 -14 -14 -37 - 24 l i 3 3 1 tel ] o
wll 0 *5 0 42
wl 2 7 -32 -37 0 0 0 0 0 0 0 0 - 7
wl 3 19 -36 -20 0 0 0 0 0 0 0 l) -S3.
wl4 4 5 -i 27 0 0 0 0 0 0 0 <1 <;
w 1 5 4 3 -IS -30 0 0 0 0 0 0 0 0 n O J u
wlO 2 5 -21 41 0 0 0 0 0 0 0 0 1 ~i
w 1 7 o i V' *1 -38 -1 0 0 0 0 0 0 0 0 -4 0
i': e J 0 -20 4 1 0 0 0 0 0 0 0 0 - 1
W i 9 28 4 5 18 0 0 0 0 0 0 0 0 - 4 5
w20 1 7 4 5 -34 0 0 0 0 0 0 0 0 - 2
v21 18 37 0 0 0 0 0 0 0 0 0 -17
w22 20 18 43 -3 -46 -8 *> o 2 2 0 0 r, <>
p to push/b to creak/ to continue:
disp/' exam/ get/ save/ set/ clear cycle do lc >5 newstart ptra
reset ru :: s t i w8ii t est
epoch 0 r s s 2.2252 7? £ ' 0.07 5 1 Sco r 0 .0000 cpr.ame p. 7
input 24 206 500 793 ' 975 9 7 5 7 93 5 00 0 0 0
actout 672 target 94 7 aelou t 60 biaout 41
acthid 438 542 279 476 448 502 o 32 - 03 502 412 507
delhid 2 2 5 0 -6 _ i o o 0 0 3
biahid J im 4 5 -38 14 -14 - 3 7 - 3 4 5 3 ; -24 10
wll 3 3 42
wi2 - -32 -3 7 0 0 0 0 0 0 0 0-7
wl3 19 -36 -20 0 0 0 0 0 0 0 0 -31
wl 4 4 5 -1 27 0 0 0 0 0 0 0 0 9
wl 5 4 3 -19 -30 0 0 0 0 0 0 0 0 18
wl6 2 5 n i u _ 41 0 0 0 0 0 0 0 0 4
wl7 34 -se -1 0 0 0 0 0 0 0 0 -40
wl8 42 -20 4 1 0 o n 0 0 0 0 0 -1
wl 9 28 1 O 18 0 0 0 o 0 0 0 0 -4 5
w20 17 '45 -34 0 0 0 0 0 0 0 0 -22
w2 i 16 2 7 o o 0 0 0 o 0 C -17
w22 20 1 8 JO o _ 6 -8 2 n O o 0 0 n 3


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain quit
reset run strain tall test
epoch 0 t s s 2 . 2854 pss 0.0602 gcor 0.0000 cpnante p.8
input 24 24 206 500 7 9 3 9 7 5 9 75 7 JJ 3 500 0 0
actout 679 target 925 delout 53 biaout 4
acthid 419 584 319 499 4 27 508 ft iL 19 421 536 431 5 06
delhid 2 2 4 0 -6 -1 O O 0 0 o
biahid -32 4 5 -38 -14 - 14 -3 7 - 0 1 C *1 0 2 3 -24 10
wll 33 42
wl 2 7 -37 0 0 0 0 0 G 0 0 - t
vl3 19 - 36 -20 0 0 0 o 0 0 0 0 -31
wl 4 4 5 -1 27 0 0 0 0 0 0 0 0 9
wl 5 43 -19 -30 0 0 0 0 0 0 0 0 18
wl 6 n r. -21 4 1 0 0 0 0 0 0 0 0 4
wl 7 34 -38 -1 0 0 0 0 0 0 0 0 -40
vl8 42 -20 41 0 0 0 0 0 0 0 0 - i
wl9 28 4 5 18 0 0 0 0 0 0 0 0 -45
w20 17 4 5 -34 0 0 0 0 0 0 0 0 _22
w21 18 37 0 0 0 0 0 0 0 0 0 -17
w22 20 18 43 __ o . 46 -8 22 22 2 0 23
p to push/b to break/' t o continue
disp/ reset exam/ run s get/ train save/ set/ clear tail test cycle do leg newstart ptra
epoch 0 tss 2.2895 pss 0.00 42 gcor 0.0006 coname p.9
input 206 24 24 206 500 793 975 975 793 500 0
actout 585 target 7 50 delout 13 biaout .1 1
acthid 404 604 346 502 423 500 230 441 576 499 527
delhid 0 0 1 0 -1 0 0 0 0 0 C
biahid 0 o i 5 -38 -14 -14 -37 -34 5 33 -24 10
wll wl2 -33 7 4 2 -22 - 3 7 0 0 0 0 0 0 0 0 !
wl 3 -19 -36 -20 0 0 0 0 0 0 0 0 - 21
wl 4 -4 5 -1 27 0 0 0 0 0 0 0 0 9
wl 5 43 -19 -30 0 0 0 0 0 0 0 0 IS
wl 6 25 -21 41 0 0 0 0 0 0 0 0 4
wl 7 _ Q £ -38 _ 1 Q 0 0 0 0 0 0 0 -40
wl 8 i n 1 *m -20 41 0 0 0 0 0 0 0 0 _ l
w 19 -28 a r IS 0 0 0 o 0 0 0 0 -4 5
w20 17 4 f- -34 0 0 0 0 0 0 0 0 _ ^ o
w21 13 3 7 0 0 0 0 0 0 0 0 0 -17
w-22 20 7 0 .1 -s 1 w _ o -46 _ 0 o 2 O 0 o o o rt


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrairi qui
reset run strain tall test
epoch 0 tss 2 # 2266 pss 0.03 71 gcor 0.0000 cpname
input 500 20C 24 24 20G 500 793 97 5 975 793 500
actout 690 target 198 delout -4 1 ti aout
acthid 399 595 3-17 182 4 39 478 2G3 479 G05 49G : 67
delhid -1 -1 -4 0 4 1 -1 _ o 0 0 - 2
biahid - 3 2 i -38 -14 -14 -37 -34 5 33 -24 10
w 1 1 i.-12 -33 1 7 42 .)*? -37 0 0 . 0 0 0 0 0 0 - 1
v. 1 3 -19 -3G -20 0 0 0 0 0 0 0 0 -31
wl 4 - 4 5 _ i ^ / 0 0 0 0 0 0 0 0 9
wl 5 43 -19 -30 0 0 0 0 0 0 0 0 18
wl 6 25 -21 ; 1 0 0 0 0 0 0 0 0 4
w 1 7 -34 -38 -1 0 0 0 0 0 0 0 0 -40
wl 8 -42 -20 41 0 0 0 0 0 0 0 0 - ]
wl 9 -28 4 5 IS 0 0 0 0 0 0 0 0 1 i
w20 17 4 5 -34 0 0 0 0 0 0 0 0
w21 18 37 0 0 0 0 0 0 0 0 0 -17
w22 20 18 4 3 -3 -46 - 8 9 9 22 2 0 9 9 to w
plO
4 1
p to push/b to break/ to continue:
disp/ exam/ get/' save/ set/ clear cycle do log
reset run strain tall test
newstar
ptrain qui
epoch 0 tss 2. 5363 pss 0.2 C97 ccor 0.0000 ccname pll
input 793 500 206 24 24 206 500 7P3 975 97 5 793
actout 690 ta rget 233 delout -97 biaout 4 1
acthid 405 560 322 4 49 4G9 452 309 519 Gii 491 622
delhid -4 -3 -9 j 11 9 -4 - 5 0 0 -5
biahid -32 45 -38 - 14 1 AT -37 -34 5 33 -24 10
wll . 3 3 42
wl 2 7 -32 - 3 7 0 0 0 0 0 0 0 0 -7
wl 3 19 -36 -20 0 0 0 0 0 0 0 0 -31
wl 4 t 0 - i 27 0 0 0 0 0 0 0 0 9
wl 5 4 3 -19 -30 0 0 0 0 3 0 0 0 18
wl 6 o s _ O 1 i 4 1 0 0 0 0 C 0 0 0 4
wl 7 34 -38 -1 0 0 0 0 C 0 0 0 -40
w 18 42 -20 4 1 0 0 0 0 0 0 0 0 -1
w: 9 28 4 5 3 8 0 0 0 0 0 0 0 0 -4 5
w20 17 H L* -34 0 0 0 0 0 0 0 0 -22
w-21 IS 37 0 0 0 (j 0 0 0 0 0 -17
w-22 20 1 o J *- _ 2 - 4 6 -8 9 9 o o 9 0 O fj


p topush/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart ptrain qui
reset run strain tall test
epoch 0 tss 2 . 9192 pss 0 .3828 gcor 0 .0000 cpname pl2
input' 9 7 5 793 500 206 24 24 20G 500 793 9 7 5 9 7 5
actout 687 target G9 aelout - 132 biaout 4 1
acthid 4 21 512 283 414 5 01 4 33 351 5 4 G CIS 4 G 7 G38 -
delhid -6 -5 -11 1 1 5 3 -6 -7 0 0 - 7
biahid - o o 45 -38 -14 - 14 - 37 - 3 4 5 33 -24 10
w 1 1 3 3 42
wl2 7 -32 -37 0 0 0 0 0 0 0 0 i
w 1 3 19 -3G -20 0 0 0 0 0 0 0 0 -31
wl4 4 5 -1 27 0 0 0 0 0 0 0 0 ij
w 1 5 43 -19 -30 0 0 0 0 0 0 0 0 18
wl 6 25 _ 21 41 0 0 0 0 0 0 0 0 4
wl 7 34 -38 -1 0 0 0 0 0 0 0 0 -40
w j 8 42 -20 41 0 0 0 0 0 0 0 0 -1
w i 9 38 4 5 18 0 0 0 0 0 0 0 0 -45
w20 17 4 5 -34 0 0 0 0 0 0 0 0 - 2 2
w21 18 3 7 0 0 0 0 0 0 0 0 0 -17
w22 20 18 43 -3 - 46 8 22 2 2 o 0 23
p to pusn/b to break/ to cont inue :
cisp/ exam/ get/ save/ set/ clear cy cle do lc reset run str ain tall test
epoch 0 tss 3.2951 pss 0.3 760 go or 0.0000 CT3 name pi 3
input 975 9 75 793 500 206 24 24 206 500 793 9 i 0
actout 682 target 69 delout .1 'iO bi aout 41
acthid 440 4 70 246 392 523 426 368 550 598 435 632
delhid -6 -5 -10 1 15 O -C - 7 0 0 - 7
biahid -32 i J *JO . 4 -14 - 3 7 - J 4 33 -24 10
wll -33 42
vl2 7 -32 -37 0 0 0 0 0 0 0 0 - 7
wl 3 -19 -3C -20 0 0 0 0 0 0 0 0 - o
wl4 -45 -1 27 0 0 0 0 0 0 0 0 9
wl5 43 -19 -30 0 0 0 0 0 0 0 0 IS
wl6 25 -21 -11 0 0 0 0 0 0 0 0 4
wl7 -34 -38 -1 0 0 0 0 0 0 0 0 -40
wlS -42 -20 41 0 0 0 0 0 0 0 0 _ i
w19 -28 45 IS 0 0 0 0 0 0 0 0 -45
w20 17 45 -34 0 0 0 0 0 0 0 0 -22
w21 IS 37 0 0 0 0 0 0 0 0 0 -17
9 IS 1 'j z <-/ s' -46 8 oo n O o 0 fx z


p to push/b to break/ to continue:
disp/ exam/ Bet/ save/ set/ clear cycle do log newstart ptrain
reset run s train tall t cst
epoch 0 tss 3.4915 pss 0.19C4 gcor 0.0000 cpname pi 4
input "S3 9 75 975 793 500 20G 24 24 20G 500 7 9 3
actout G76 target 233 delout -57 biaout 41
acthid 455 449 224 390 52G 435 3 53 529 5G7 4 0 7 G04
delhid -4 -3 7 1 11 2 -5 5 0 0 5
biahid -32 4 0 -38 -14 -14 -37 - 34 5 33 -24 10
wll -33 4 2
wl 2 1 -32 -37 0 0 0 0 0 0 0 0 - 7
wl 3 -19 -3G -20 0 0 0 0 0 0 0 0 -21
w] 4 -45 _ i o - 0 V 0 0 0 0 0 0 9
v 1 5 43 -19 -30 0 0 0 0 0 0 0 0 18
w 1 G 25 -21 41 0 0 0 0 0 0 0 0 1
wl 7 -34 -38 _ i 0 0 0 0 0 0 0 0 -40
wlS -42 -20 4 1 0 0 0 0 0 0 0 0 _ 1
wl 9 -28 45 18 0 0 0 0 0 0 c 0 -45
w20 17 *n 0 -34 0 0 0 0 0 0 0 0 _ O
w21 18 37 0 0 0 0 0 0 0 0 0 -17
w22 20 18 43 -3 -4G -8 22 2 2 o 0 3
p to push/b to break / < c : > to ( continue
disp/ exam/ get/ save/ set/ clear cycle do lo '£ news tart ptrain
reset run St rain tail test
epoch 0 tss 3.521 7 pss 0.0302 gccr 0.0000 cpnamc pi 5
input 0 00 793 975 9 75 793 500 2 00 24 24 20G 00
actout e 71 targe t 498 delout -38 bi aout 41
acthid 461 459 223 408 510 456 313 491 538 392 564
delhid -1 -1 -2 0 4 i -2 0 0 _ 2
biahid - 0 V <. i c o 8 14 -14 -37 - -i 0 uO -24 10
wl 1 -33 A O
w 12 i -32 -37 0 0 0 0 0 0 0 0 -7
w 13 -19 -36 -20 0 0 0 0 0 0 0 0 -31
w ] 4 -4 5 -1 2 7 0 0 0 0 0 0 0 0 9
wl5 43 -19 -30 0 0 0 0 0 0 0 0 18
wl 6 25 -21 41 0 O' 0 0 0 0 0 0 **
wl 7 -34 -38 -1 0 0 0 0 0 0 0 0 -40
wl 3 -42 -20 41 0 0 0 0 0 0 0 0 _ i
wl 9 -28 4 o 18 0 0 0 0 0 0 0 0 - 4 5
w20 17 15 -34 0 0 0 0 0 0 0 0 -22
w21 18 2 7 0 0 0 0 o. 0 0 0 "0 -17
,:o o 20 1 Z 4 3 3 _ 0 n2 O r> O o 2 3
qui t
qui t


p to push/b to break/ to continue:
disp/ exam/ get/ save/ set/ clear cycle do log newstart
reset run strain tall test
epoch 0 tss 3.3302 pss 0.0085 gcor 0.0000 cpname
input 20G 500 7S3 97 5 9 7 5 793 tOO 20G 24 2-3 206
actout G70 target 763 delout 20 b i a o u t
acthid 454 494 243 441 480 481 267 451 521 397 528
delhid 1 0 1 0 -2 0 0 1 0 0 1
bialiid -32 45 -38 -14 -14 -37 -34 5 33 24 10
wll no \ n %J <3 *J L.
wl2 7 -32 -27 *0 0 0 0 0 0 0 0 -7
wl3 19 -36 -20 0 0 0 0 0 0 0 0 -31
wl4 45 -i 27 0 0 0 0 0 0 0 0 9
w 1 5 43 -19 -30 0 0 0 0 0 0 0 0 IS
wlG 25 -21 41 0 0 0 0 0 0 0 0 4
k1 7 34 -38 -1 0 0 0 0 0 0 0 0 -40
wie 42 -20 4 1 0 0 0 0 0 0 0 0 -1
wl 9 28 45 18 0 0 0 0 0 0 0 0 -45
w20 17 45 -34 0 0 0 0 0 0 0 0 -22
w21 18 37 0 0 0 0 0 0 0 0 0 -17
w22 20 18 43 -3 -46 _ g O 0 o o 2 0 2 3
p to push/b to break/(cr) to cor; tir.ue:
disp/ exam/ get/ save/ set/ clear cycle i cc log newstart
reset run strain tali test
epoch 0 tss 3.5942 ps s 0.0639 gcc.r 0.0000
input 24 20G 500 79 3 9 7 5 £75 793 500 206 24 2 4
actout G74 target 92 7 delout 5 5 biaout
acthid 438 542 279 47 6 448 502 232 424 524 J 20 510
delhid 2 2 4 0 -6 _ 2 o o 0 0 3
slabid 32 -rj j 8 1 4 -14 C a 1 u 1 9 n n - 24 10
i.-l 1 33 4 2
wl2 7 -32 -37 0 0 0 0 0 0 0 0 -7
*5 i> o 19 -36 -20 0 0 0 0 0 0 0 0 -21
i.-14 45 -1 27 0 0 0 0 0 0 0 0 9
i.-15 43 -19 -30 0 0 0 0 0 0 0 0 18
wl G 25 -21 41 0 0 0 0 0 0 0 0 4
1.-17 34 -38 -1 0 0 0 0 0 0 0 0 -40
i.-18 42-20 41 0 0 0 0 0 0 0 0 -1
wl 9 28 45 16 0 0 0 0 0 0 0 0 -45
i.-20 17 45-34 0 0 0 0 0 0 0 0 -22
1-21 IS 27 v 0 r. 0 0 0 0 0 0 -17
w-22 20 18 48 -2 -4 6 _ o r>o o o 2 0 n 3
ptrain qui
p 16
11
qui
P- '