ml on medical images - keras

[avi.e.gross at gmail.com]

Nati,

You ask what the problem is in that code. I say there I absolutely NO PROBLEM for me in that code.

Do you know why?

Because even if I want to copy it and make sure I have all the modules it needs, I have no access to the files it opens, and no idea what the output of the images pumped in is supposed to be and so on.

So since I won't try it, it stops being my problem.

Now if you want to know if someone spied any obvious error in the code, who knows, someone might. 

But guess who has access to all the code and files and anything else in your environment? You!

So why not use your interactive python interpreter and pause between parts of the code and insert requests to see what things look like, which is what programmers have been doing for decades, or use some debugger?

Check what is being done and if something fails, trace back as to what it was being given and see if that is what the manual page suggests you should be giving it and so on.

You cannot expect multiple people to keep doing the work for you and especially if they keep telling you they need more info.

I know very little about you and what tasks you have agreed to do but suggest that it may end up being a lot more work than you anticipated given how many kinds of pseudo-questions you seem to have.

What is wrong with your code is that someone else did not write it specifically for the purpose you want. My guess is that you copy lots of code from libraries or the internet and want to adapt it without necessarily understanding if it fits your needs or where it needs to be tweaked.

If I am wrong, I apologize. But if you want help here, or in other forums, consider changing your approach and consider what you would want if anyone else asked you to help them.

What you keep not wanting to do is supply a fairly simple example of inputs and outputs and error messages. Now in this case, if your inputs are images and machine learning algorithms are going to  output best guesses about features such as what animal is in the picture, it may indeed be harder to give serious details. 

When I see a question I can answer, I may chime in but for now, this process is too frustrating.

Avi

-----Original Message-----
From: Python-list <python-list-bounces+avi.e.gross=gmail.com at python.org> On Behalf Of ??? ????
Sent: Monday, July 11, 2022 3:26 AM
To: python-list at python.org
Subject: Fwd: ml on medical images - keras

---------- Forwarded message ---------
מאת: נתי שטרן <nsh531 at gmail.com>
‪Date: יום א׳, 10 ביולי 2022, 13:01‬
Subject: Re: ml on medical images - keras
To: Neuroimaging analysis in Python <neuroimaging at python.org>


p.s. all the pictures are in PNG FORMAT

‫בתאריך יום א׳, 10 ביולי 2022 ב-13:01 מאת נתי שטרן <‪nsh531 at gmail.com‬‏>:‬

> What's the problem on this code:
>
> import os
> from pickletools import float8, uint8
> from PIL import Image
>
> import numpy as np
> import tensorflow as tf
> from tensorflow import keras
> from tensorflow.keras import layers
> inputs=[]
> for i in os.listdir("c:/inetpub/wwwroot/out"):
>     for j in os.listdir("c:/inetpub/wwwroot/out/"+i+"/"):
>             A=Image.open("c:/inetpub/wwwroot/out/"+i+"/"+j)
>             from matplotlib import pyplot as plt
>
>             filename = 'image-test'
>
>
>           #  img = ( filename + '.png' )
>             x=np.array(A,dtype=np.shape(A))
>             inputs.append(x)
>
> simple_rnn = tf.keras.layers.SimpleRNN(4) #np.int output = 
> simple_rnn(inputs=np.array(inputs,dtype=np.ndarray(260, 730, 4)))  # 
> The output has shape `[32, 4]`.
>
> simple_rnn = tf.keras.layers.SimpleRNN(
>     4, return_sequences=True, return_state=True)
>
> # whole_sequence_output has shape `[32, 10, 4]`.
> # final_state has shape `[32, 4]`.
> whole_sequence_output, final_state = simple_rnn(inputs)
>
> --
> <https://netanel.ml>
>


--
<https://netanel.ml>
--
https://mail.python.org/mailman/listinfo/python-list