Hello, everyone! My name’s Sophie and I’m an intern at Maplesoft. @Samir Khan asked me to develop a couple of demonstration applications using the DeepLearning package - my work is featured on the Application Center

• Prediction of malignant/benign of breast mass with DNN classifier

• Predicting the burnt area of a forest fires with DNN Regressor

I thought I’d describe two critical commands used in the applications – DNNClassifier() and DNNRegressor().

The DNNClassifier calls tf.estimator.DNNClassifier from the Tensorflow Python API. This command builds a feedforward multilayer neural network that is trained with a set of labeled data in order to perform classification on similar, unlabeled data.

Dataset used for training and validating the classifier has the type DataFrame in Maple. In the Prediction of malignant/benign of breast mass example, the training set is a DataFrame with 32 columns in total, with column labels: “ID Number”, “Diagnosis”, “radius”, “texture”, etc. Note that labeling the columns of the dataset is mandatory, as later the neural network needs to identify which feature column corresponds to which list of values.

Feature columns are what come between the raw input data and the classifier model; they are required by Tensorflow to specify how the input data should be transformed before given to the model. Maple now supports three types of Feature Columns, including:

• NumericColumn that represents real, numerical figure,
• CategoricalColumn that denotes categorical(ordinal) data
• BucketizedColumn that organizes continuous data into a discrete number buckets with specified boundaries.

In this application, the input data consists of 30 real, numeric values that represents physical traits of a cell nucleus computed from a digitized image of the breast mass. We create a list of NumericColumns by calling

with(DeepLearning):
fc := [seq(NumericColumn(u,shape=[1]), u in cols[3..])]:

where cols is a list of column labels and shape[1] indicates that each data input is just a single numeric value.

When we create a DNNClassifier, we need to specify the feature columns (input layer), the architecture of the neural network (hidden layers) and the number of classes (output layer). Recall that the DNNClassifier builds a feedforward multilayer neural network, hence when we call the function, we need to indicate how many hidden layers we want and how many nodes there should be on each of the layer. This is done by passing a list of non-negative integers as the parameter hidden_units when we call the function. In the example, we did:

classifier := DNNClassifier(fc, hidden_units=[20,40,20],num_classes=2):

where we set 3 hidden layer each with 20, 40, 20 nodes respectively. In addition, there are 30 input nodes (i.e. the number of feature columns) and 1 output node (i.e. binary classification). The diagram below illustrates a simpler example with an input layer with 3 nodes, 2 hidden layers with 7, 5 nodes and an output layer with 1 node.

(Created using NN-SVG by https://github.com/zfrenchee/NN-SVG)

After we built the model, we can train it by calling

classifier:-Train(train_data[3..32], train_data[2], steps = 256, num_epochs = 3, shuffle = true):

where we

1. Give the training data (train_data[3..32]) and the corresponding labels (train_data[2]) to the model.
2. Specified that the entire dataset will be passed to the model for three times and each iteration has 256 steps.
3. Specified that data batches for training will be created by randomly shuffling the tensors.

Now the training process is complete, we can use the validation set to evaluate the effectiveness of our model.

classifier:-Evaluate(test_data[3..32],test_data[2], steps = 32);

The output indicates an accuracy of ~92.11% in this case. There are more indices like accuracy_basline, auc, average_loss that help us decide if we need to modify the architecture for better performance.

We then build a predictor function that takes an arbitrary set of measurements as a DataSeries and returns a prediction generated by the trained DNN classifier.

predictor := proc (ds) classifier:-Predict(Transpose(DataFrame(ds)), num_epochs = 1, shuffle = false)[1] end proc;

Now we can pass a DataSeries with 30 labeled rows to the predictor: (Recall the cols is a list of the column names)

ds := DataSeries([11.49, 14.59, 73.99, 404.9, 0.1046, 8.23E-02, 5.31E-02, 1.97E-02, 0.1779, 6.57E-02, 0.2034, 1.166, 1.567, 14.34, 4.96E-03, 2.11E-02, 4.16E-02, 8.04E-03, 1.84E-02, 3.61E-03, 12.4, 21.9, 82.04, 467.6, 0.1352, 0.201, 0.2596, 7.43E-02, 0.2941, 9.18E-02], labels = cols[3..]); 
predictor(ds);

The output indicates that the probability of this data being a class _id [0] is ~90.79%. In other words, according to our model, the probability of this breast mass cell being benign is ~90.79%.

The use of the DNNRegressor is very similar (almost identical) to that of the Classifier, the only significant difference is that while the Classifier predicts discrete labels as classes, the Regressor predicts a continuous qualitative result with the provided data (Note that CategoricalColumn is still applicable). For more details about the basic usage of the DNNRegressor, please refer to Predicting the burnt area of a forest fires with DNN Regressor.

An error message during opening a Maple file is seen:

ibb.co/hqOOkJ

I put the file link in below. The Maple file has been saved as Maple 2016.

https://files.fm/u/jpc3k5s4

I will be grateful if you can recover the file. Thanks

I would like to plot a graph whose legend has a Greek character with a numeric subscript. Below, there is a small example, where I show what I am trying to do:

restart:

# Simplified example, which is not working:

omega0:= 10:
plot(omega^2/omega0, omega = 1..1.5,
legend = [sprintf("%s = %.1f",`ω`[0],omega0)]);

# The character I want to be plotted inside the legend:

`ω`[0];

I suppose the problem is that " `ω[0]` is not considered a string and, therefore, I can not call it in the legend with a "%s". I do not know how to make it work, though. Does anyone know how to do so?

From help it says

"For any variable used within a procedure without being explicitly
mentioned in a local localSequence; or global globalSequence; the
following rules are used to determine whether it is local or global:

The variable is searched for amongst the locals and globals (explicit or implicit)
in surrounding procedures, starting with the innermost.  If the name is
encountered as a parameter, local variable, or global variable of such
a surrounding procedure, that is what it refers to."

--------------------------------------------

So it seems if I do not use explicit "global" on a variable, Maple can figure
if it is global or not using the above rules. But when I use explicit "global"
on a variable, Maple did not seem to do the same thing. Here is an example
 

restart;
foo:=proc()
  local a, inner_proc;

  inner_proc := proc()
     local b;
     b   := ithprime(10);
     a   := b; #this assiged the global (to this proc)
               #variable, which is "a" correctly
  end proc;

  inner_proc();

  return(a);
end proc;

calling foo() gives 29.

But since the variable "a" in foo() is a global with respect to the inner_proc() (based on what the above help page seems to say), then why this does not work

restart;
foo:=proc()
  local a, inner_proc;

  inner_proc := proc()
     local b;
     global a;
     b   := ithprime(10);
     a   := b;                
  end proc;

  inner_proc();

  return(a);
end proc;

Now foo() returns "a". So Maple did not assign 29 to the global "a", (global to the inner_proc).

This for me makes little sense. Is global in Maple means the outermost scope only, skipping everything in between?

I thought global means any variable outside the proc itself. So if the proc() was inside another proc(), then the variables in the outer proc are global to the inner proc, even if they are declared local to the outer proc.

Why did Maple not do the assignment when I explicitly declare "a" to be global in the inner proc?

If you type ithprime (10^7) and hit return, the answer 179424673 is returned instantaneously, but if you do ithprime(10^8) there is no answer in ~ one hour plus (and maybe never), it just says “evaluating” and no answer comes. I have done calculations in Maple where the answer was a prime with ~18,000 digits yet cannot get beyond ithprime(10^7) when writing a code that needs to work up the prime list, looking for a particular one. Has anyone else encountered this apparent limitation in Maple?

Is there a way around it? 

I'm having a hard time trying to understand and use the various fitting functions.

For starters, if I copy and paste the code from the example on the Maple support Help page for NonlinearFit(), I'm not able to successfully execute the code. It gives the error:

Error, (in Statistics:-NonlinearFit) complex value encountered

I've attached my Maple file: testing_nonlinearfit.mw

Hello,

     I'm having trouble simplifying this square root:

assume(p::positive):
sqrt(sqrt(p^2+1)-1)*sqrt(sqrt(p^2+1)+1):
expand(%);

I expect it to give just p but instead it returns the full, unsimplifed expression.

This appears as part of a larger expression, and if it fails to simplify, terms won't cancel and the expression is much longer than it needs to be.

Thanks!

Mukhametshina Liya

Games with pseudo-fractals
 

Homothety_Fractals.mw

I would like to define a fourth-order tensor (3x3x3x3) with zeros, for then defining some non-null terms individually. For that, I'm currently doing:

for it1 from 1 to 3 do
    for it2 from 1 to 3 do
        for it3 from 1 to 3 do
            for it4 from 1 to 3 do
                H[it1,it2,it3,it4]:= 0;
            end do:
        end do:
    end do:
end do:

H[1,1,1,1]:= value1:
H[1,2,2,1]:= value2:
# etc.

However, I suppose there is probably a more efficient manner to do do, maybe using Physics package. Does anyone know to do that?

Hello,

I have a very simple problem. When Maple displays long outputs I can only see a part of them. Here there is an example

https://www.dropbox.com/s/ymp1vdsg80ewu1s/Untitled.jpeg?dl=0

On my previous versions of Maple I had a slider on the bottom of the page. How can I activate it in Maple 2016?

Thanks, Nicola

Aleksandrov Denis,
7th grade
secondary school #57 of Kazan

_ANIMATED_PICTURE_ON_THE_COORDINATE_PLANE_Aleksandrov_D..mw

Why does Maple dsolve give this strange error from dsolve? I do not see that the input is wrong

restart;
F := x * ( y(x) + x*sqrt(x*y(x)) + sqrt(x^3*y(x)) );
ode:= diff(y(x),x) = F;
dsolve(ode,y(x));

I do not see where the input is invalid. I stared at it for 5 minutes.

Mathematica can solve this as follows

ode = y'[x] == x ( y[x] + x Sqrt[x y[x]] + Sqrt[x^3 y[x]]  )
DSolve[ode, y[x], x]

Did I type something wrong in Maple?

Physics:-Version();
    2018, June 12, 1:40 hours, MapleCloud version: 60

I want z  to becomes abs(z) then later on, remove abs() and obtain z back in the same form it was.

I found that Maple changes z when I put it inside abs. Like this

expr:=(-y^2+1);
abs(expr);
op(1,%);

So instead of 1-y^2, I end up now with  y^2-1. I want to keep the same expression I started with.

Mathematica does not do this:

expr = 1 - y^2
Abs[expr]
%[[1]]

Is there a way or option to tell Maple not change the expression when I put it inside abs()?

This complicates what I am trying to do in Maple.

thank you

Hello,

I've tried many combinations of assume(), simplify(), expand(), and evalc(), and I simply can't figure out how to make Maple evaluate my symbolic expression as a purely real valued expression. I know that my expression must be able to be expressed solely in real-valued terms since any number times its complex conjugate is real-valued. What function do I use to accomplish this? Thanks for your consideration. I've attached the code in question below.

Ian

testing_real_simplification.mw

restart;
v := 6000;

# Define the Fresnel reflection coefficient, rp
eta0 := 4; eta1 := n1+I*kappa1;
theta0 := (1/180)*ang*Pi; theta1 := arcsin(eta0*sin(theta0)/eta1);
xi0 := eta0*cos(theta0); xi1 := eta1*cos(theta1);
rp := (-eta0^2*xi1+eta1^2*xi0)/(eta0^2*xi1+eta1^2*xi0);

# Define the complex conjugate of the Fresnel reflection coefficient, rpCon
eta1Con := evalc(conjugate(eta1));
theta1Con := arcsin(eta0*sin(theta0)/eta1Con);
xi1Con := eta1Con*cos(theta1Con);
rpCon := (-eta0^2*xi1Con+eta1Con^2*xi0)/(eta0^2*xi1Con+eta1Con^2*xi0);

# Take the product of rp and its complex conjugate, rpCon. This should necessarily evaluate to a real-valued quantity.
prod := evalc(rp*rpCon);