Convert undirected graph unweighted graph to a weighted graph even though all the weights are 1

as i want to check how my function works on weighted graph as I know it works unweighted

Can someone give me an example for using

linear regressor function from deep learning package

By passing one dependent variable say Y

And a set of independent variables say X

As a matrix say 

Can we be able to split data into train and test and use this linear regressor command on train and do validation

As I am not able to see some examples on the implementations using command kind help 

Hope the resolution of the images which come via Plot3d and plot commands are atleast 300 dpi

If not how to make them

Convert a table in a form 

output≔table([(2,4)=["O-H",0.97234632],(1,2)=["O-O",1.44940000],(1,3)=["O-H",0.97232285]])

Table can have more elements in a similar form at (1,2) and (2,1) position 

In the new matrix converted

We should get a square matrix

With at (2,4) and (4,2) position 

0.97234632

Similarly at (1,2) and (2,1) position 1.44940000

And so on 

Remember a square matrix and it is symmetric 

Fix a plot output image size to a certain big size instead of going to each plot and then trying to maginify it where output runs through pages.