🔒 Bad File Descriptor problem in Python

[lean lava]

I am trying to write to a text file in Python. The name of the file is testfile.txt. Here is what I tried:

f = open('testfile.txt', 'r')

print(f.read())
Hello out there.

f.close()
f = open('testfile.txt', 'a')
f.writable()
True
f.write('Right back at ya!')
17
f.close()
OSError: [Errno 9] Bad file descriptor

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "<stdin>", line 1, in <module>
OSError: [Errno 9] Bad file descriptor

Even thought it looks like it is writing to the file, it isn't. What am I doing wrong? Thanks in advance.

[stable widget]

@lean lava

Python help channel opened

Remember to:

• Ask your Python question, not if you can ask or if there's an expert who can help.
• Show a code sample as text (rather than a screenshot) and the error message, if you've got one.
• Explain what you expect to happen and what actually happens.

:warning: Do not pip install anything that isn't related to your question, especially if asked to over DMs.

Closes after a period of inactivity, or when you send !close.

[tepid marlin]

def backward(self, dy):
# Backward pass through the RNN
batch_size, seq_len, output_dim = dy.shape
dWy = np.zeros_like(self.Wy)
dby = np.zeros_like(self.by)
dWh = np.zeros_like(self.Wh)
dbh = np.zeros_like(self.bh)
dWx = np.zeros_like(self.Wx)

    dh_next = np.zeros_like(self.h)
    
    for t in reversed(range(seq_len)):
        dy_t = dy[:, t, :]
        dWy += np.dot(self.h_stack[:, t, :].T, dy_t)
        dby += np.sum(dy_t, axis=0, keepdims=True)
        
        dh_t = np.dot(dy_t, self.Wy.T) + dh_next
        dh_tanh = (1 - self.h_stack[:, t, :] ** 2) * dh_t
        
        dWx += np.dot(self.x[:, t, :].T, dh_tanh)
        dbh += np.sum(dh_tanh, axis=0, keepdims=True)
        dWh += np.dot(self.h_prev.T, dh_tanh)
        
        dh_next = np.dot(dh_tanh, self.Wh.T)

    This is the backpropagation for rnn. In weight adjusting we use dot product but for bias we use summation, please explain this

[tender steppe]

I tried this on my own system and it worked as expected. Does your program contain any other logic that may interfere with the file I/O?

[tender steppe]

ideally you should create your own help post (#❓｜how-to-get-help) 👍

[lean lava]

Thank you for responding. But no, there is nothing else. What I posted was the interaction from the command prompt. There is nothing more going on that what you see. I am new to this - my first time posting. I look at how to create a help post. Thanks again.

[lean lava]

So I looked at the instructions on how to create a help post and thought that is what I did. What did I do wrong?

[tender steppe]

sorry, i was responding to the other person who hijacked your post 😅

[lean lava]

No problem. Any other thoughts on this. It seems very strange to me.

[tender steppe]

Does the program work if you run it from a file (eg main.py)? I am not 100% sure why you have this issue, but it might be exclusive to the Python REPL (command prompt) because the f deconstructor could have been called somehow before you called f.close.

[lean lava]

The same thing happens when I run it from a file. I am using VSCode.

[prisma ridge]

Show the full code you ran and the full traceback

Remember to use code blocks

!code

[stable widget]

Formatting code on Discord

Here's how to format Python code on Discord:

```py
print('Hello world!')
```

These are backticks, not quotes. Check this out if you can't find the backtick key.

For long code samples, you can use our pastebin.

[lean lava]

Here is all the code there is in the file:
with open('testfile.txt', 'w') as f:
f.write("right back at ya")
There is no more to it than that. Here is message:
OSError: [Errno 9] Bad file descriptor

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "c:\Users\rwalt\Documents\pcode\hello\writing.py", line 1, in <module>
with open('testfile.txt', 'w') as f:
OSError: [Errno 9] Bad file descriptor

It seems like this in my previous message I just the version where I interacted directly with the command prompt.

[lean lava]

On the other hand, if I run the following file:

with open('testfile.txt', 'r') as f:
print(f.read())

I get:

Hello out there.

which is exactly what is in the file. So would it work for read and not for write?

[prisma ridge]

That's weird

[lean lava]

Yep. Any other thoughts? I am completely mystified.

[prisma ridge]

Are you sure there's not more traceback, there's a "During handling of the above exception" bit

So I suspect there's more output you're not showing

[lean lava]

I promise there is no more than what I showed. After the "During handling of the above exception" the rest is:

Traceback (most recent call last):
File "c:\Users\rwalt\Documents\pcode\hello\writing.py", line 1, in <module>
with open('testfile.txt', 'w') as f:
OSError: [Errno 9] Bad file descriptor

There is no more than that.

[prisma ridge]

There should be more before that

Remember to use code blocks

[lean lava]

Really, there is no more. I copied it right off the screen.

[lean lava]

Really, there is no more than that. I copied it right off the screen.

[prisma ridge]

Can you send a screenshot?

[lean lava]

OK. Hold on. How do I send the screen shot on this. I have never used this before.

[lean lava]

I'm sorry, but how do I send the screen shot? I have never used this before.

[lean lava]

OK. Here is the screen shot.

[lean lava]

This is the screen shot from VSCode.

[prisma ridge]

There's bizarre

[lean lava]

Any other thoughts?

[lean lava]

And here is the screen shot of the same interaction from the command prompt. I even checked to see if the file was writable and it appears to be so.

[cloud tartan]

@lean lava When you open a file using with open, the end of the with closes the file for you. When you go f.close() after that, the file is already closed. The OS file descriptor associated with the file is no longer valid, since it isn't associated witrh an open file.

[tender steppe]

what you describe is not happening in the code.

[tender steppe]

i read on-line that sometimes Windows Defender exclusions can cause a similar error. Do you have an anti-virus exclusion for ~/Documents or pcode for example? 🤔 it might be some kind of permission issue

[lean lava]

In the screenshot with the code, I did not call f.close(). I only used the 'with' statement. What you see in the screen shot is really all there is. I did not use the with statement in command prompt interaction and got the same error. So the use of the 'with' is not the problem.

[lean lava]

I don't know, but I will check it out. I am running Windows Defender. Thanks.

[cloud tartan]

As you say. Sorry.

[stable widget]

Python help channel closed

This help channel has been closed and it's no longer possible to send messages here. If your question wasn't answered, feel free to create a new post in #1035199133436354600. To maximize your chances of getting a response, check out this guide on asking good questions.