[Feature Request] Run chunks akin to PyCharm Scientific Mode

[j9ac9k]

Environment data

• VS Code version: 1.27.1
• Extension version (available under the Extensions sidebar): 2018.8.0
• OS and version: macOS 10.13.6
• Python version (& distribution if applicable, e.g. Anaconda): Anaconda, pyenv 3.7 (any)
• Type of virtual environment used (N/A | venv | virtualenv | conda | ...): conda virtualenv pyenv
• Relevant/affected Python packages and their versions: any

Actual behavior

Code executes as it should

Expected Requested behavior

PyCharm a year ago unveiled a "Scientific mode" feature, which I think visual studio would benefit from having a feature similar to it. It allows for developers to easily mark chunks of code to execute and then to interact with the live interpreter.

Here is a screenshot of an example from within PyCharm

By adding #%% the user initiates effectively a breakline, where they can run a chunk of code.

The use case here is primarily for pandas, or numpy operations where you need to test out that you're calling the right arguments, or have the right method chaining, which takes a little while to get working right. Often complicated merge operations can take multiple iterations to work as intended. Having the IPython interpreter active on the bottom also makes it easy to query various objects to see what their shape/values are.

It would be great is visual studio code had a feature similar to this.

Steps to reproduce:

1. NA

Logs

Output for Python in the Output panel (View→Output, change the drop-down the upper-right of the Output panel to Python)

NA

Output from Console under the Developer Tools panel (toggle Developer Tools on under Help)

NA

[d3r3kk]

If you run your script in the debugger, and break in a place (in the script above, on lines 95 & 109) you should be able to interrogate local objects in the interactive debugger window in VSCode with the Python extension today.

Is this what you are looking to do?

[j9ac9k]

Right, should have addressed the possibility of using breakpoints (which is possible) but there are a few issues.

1. The debug console does not offer any form of autocomplete that IPython offers (which I suppose that's it's own issue .... use IPython when it's visible within a python environment). Looks relevant to #727
2. Generally speaking, running code through a debugger is much slower. Some of the larger pandas operations can take a while.
3. After the breakpoint, if I make changes to the underlying python file, (to my knowledge) it won't be recognized until I stop/restart the debugger.

EDIT: looks like my feature request is very similar to a #2302

[brettcannon]

#2462 is to add completions to the debug console.

And yes, I would say this is a duplicate of #2302.