Bug: The variable explorer doesn't show the shape of np arrays in the "count" column

[jay-pee]

Bug: The variable explorer doesn't show the shape of np arrays in the "count" column

Steps to cause the bug to occur

1. Create two numpy arrays:

import numpy as np
x = np.array([1,2,3])
y = np.array([[1,2,3],[1,2,3]])

2. Look at the variable explorer in the interactive window

Actual behavior

The "count" column shows only the number of rows I assume.

Expected behavior

I saw in a YouTube video that the "count" column should actually show the shape (x.shape)

I use a lot of linear algebra calculations with multi-dimensional arrays, so for me it is quite important to see the shape of an array at a glance. Workaround for me is using the iPython REPL to get the shape, which is clunky.

Your Jupyter and/or Python environment

Please provide as much info as you readily know

• Jupyter server running: Local
• Extension version: 2020.3.69010
• VS Code version: 1.43.1
• Setting python.jediEnabled: false
• Python and/or Anaconda version: 3.7.7
• OS: Windows
• Virtual environment: N/A

Microsoft Data Science for VS Code Engineering Team: @rchiodo, @IanMatthewHuff, @DavidKutu, @DonJayamanne, @greazer

[jmew]

Thanks for logging the issue @jay-pee. I was able to verify it on our end too so i think something changed since our december build of the extension.

Might be a regression? cc @rchiodo

[jmew]

nevermind, just looked through our previous issues. It's a duplicate of microsoft/vscode-python#10140. The explanation for why we can only show one dimension is below:

"We switched computing the value of the variable to use the jupyter ? command instead of running python code. This output doesn't seem to include the shape for ndarrays.

We did this so we could start supporting other kernels besides python."

[jay-pee]

Thank you for your fast answer.

If its not a bug then see it as a feature request ;) In my opinion the "count" column is quite meaningless for numpy arrays and maybe also confusing in the current state.

[greazer]

While this is a regression, it's not quite meeting the bar for must fix at this time.

[jaewooklee93]

I think this feature is really important for many data scientists.

[IanMatthewHuff]

Validated.