Yannick Funk comments

Results 17 comments of


                                            Yannick Funk

.mean() on ArrayView1<u64>: panicked at 'attempt to add with overflow'

@bluss yes, it should be just the float without the fractional part, just the behavior the mean function has when calculating the mean of few u64's (not 1000 like in...

.mean() on ArrayView1<u64>: panicked at 'attempt to add with overflow'

@xd009642 would casting of the sum results if they exceed u64 be an option?

.mean() on ArrayView1<u64>: panicked at 'attempt to add with overflow'

@xd009642 I'm using mean_axis at the moment to calculate the means of groups of `u64` values along an axis. Can you imagine any workaround at the moment to not encounter...

.mean() on ArrayView1<u64>: panicked at 'attempt to add with overflow'

@bluss thanks! So maybe the first step in this issue should be adding the mentioned panic case to the functions documentation: https://github.com/rust-ndarray/ndarray/blob/4d9641db085e64729d52050ea0c24e4a541629a7/src/numeric/impl_numeric.rs#L70 What do you think?

.mean() on ArrayView1<u64>: panicked at 'attempt to add with overflow'

I personally would volunteer to implement this, but I never contributed here so I might need some Information where to start when making such a change. @jturner314

[FEATURE] Provide better Error messages, with a priority on REPLs

To achieve this we would need to think of how to accomplish the following: - [ ] Distinguish user written python code from pyscript internal python code (e.g. calls to...

[FEATURE] Provide better Error messages, with a priority on REPLs

@marimeireles Yeah the output comes directly from pyodide, my intention here was to parse this output and hide the pyodide internals from the user. That's what I meant with >...

Neural networks doesn't function in pyscript

If you still want NNs to be working, try sklearn! It works with pyscript and supports MLPs (Fully Connected neural Networks): https://scikit-learn.org/stable/modules/generated/sklearn.neural_network.MLPClassifier.html?highlight=mlp#sklearn.neural_network.MLPClassifier

multiprocessing module not working as expected - _multiprocessing missing (not supported?!)

Thank you for submitting the issue! As stated in the issue linked by you (https://github.com/pyodide/pyodide/issues/1603) > multiprocessing won't work in Pyodide because one cannot create processes (and threads for now)....

Streaming Audio is choppy

Is there any progress here? This feature would be much appreciated