Run code in codespace

[npr99]

GitHub has a new option to run code in a virtual environment called codespace. This sounds like a good idea but - when I initially launch the codespace the environment is not setup. I had expected the environment to be based on the .yml file. But it looks like I need to make a .dev file.

Following these instructions:
https://docs.github.com/en/codespaces/setting-up-your-project-for-codespaces/adding-a-dev-container-configuration/setting-up-your-python-project-for-codespaces

[npr99]

I use 3.10.9 in the last environment but I do not know if that is bookworm, buster or bullseye. Looks like these have something to do with the docker image. I will try the default 3.12 bullseye

[npr99]

following instructions and selecting coverage.py.... not sure what this does

[npr99]

add python packages to devcontainer.json github codespace
https://stackoverflow.com/questions/74648852/how-to-install-python-package-on-github-codespaces-without-having-to-rebuild-the

looks like I need a requirements.txt file

https://www.reddit.com/r/learnpython/comments/xvlpdz/why_do_people_provide_a_requirementstxt_or/

[npr99]

Tip: You may occasionally want to perform a full rebuild to clear your cache and rebuild your container with fresh images. For more information, see

[npr99]

looks like something is happening. I made a requirements files with the primary conda packages. I did not include the pip packages

pandas=1.3.5
geopandas=0.12.2
requests=2.27.1
seaborn=0.11.2
scooby=0.5.12
jupyter=1.0.0
contextily=1.3.0

[npr99]

the codespace still wanted me to start a new kernel.

[npr99]

updated the requirements files to include ipywidgets==7.7.1... that was the error I just got. Will see if rebuilding the codespace shows that the requirments file is doing something.

[npr99]

still have issue....using this example to see if that fixes some of the problems.
https://github.com/github/codespaces-jupyter/blob/main/.devcontainer/devcontainer.json

[npr99]

The example from codespaces-jupyter works from their repo.... I did not have to pick the kernel

[npr99]

test comment getting error message saying can not save comment

[npr99]

ipywidgets==7.7.1
matplotlib==3.7.0
numpy==1.24.2
pandas=1.3.5
geopandas=0.12.2
requests=2.27.1
seaborn=0.11.2
scooby=0.5.12
contextily=1.3.0

[npr99]

just noticed that requirements uses == not =

[npr99]

this time it seems to have selected python 3.10.13 - it took a second but it seems to have done that. But it still can not find the ipywidgets... it seems to have installed the requirements file automatically. I just had to run the first block of the main notebook and wait. The terminal showed that the installation was happening.

[npr99]

the above error seems to have negatively impacted the notebook. I tried to run the command 2+2 but the code block just ran and ran

[npr99]

rebuild the codespace... took about minutes. Automatically ran:

[npr99]

seems to be working - need to add
fpdf2==2.5.2
wget==3.2
to requirments. But since they are pip installed not sure if it will work. I was able to install both in the terminal.

[npr99]

also missing folium
in terminal conda install folium
install folium
Name Version Build Channel
folium 0.14.0 py312h06a4308_0

but getting error that it is not installed.
added to the requirements file and rebuilding

[npr99]

rebuild seems to have worked.
code is working found a bug in the directory_design file

needed to add
outputfolder = "OutputData"

Make directory to save output
if not os.path.exists(outputfolder):
os.mkdir(outputfolder)

[npr99]

issue with the family variable having missing values?
added mechatroner.rainbow-csv
need to add to codespace devcontainer
Issue replicates outside of codespace... need to address in another issue

[npr99]

codespace appears to be working.

got a
RemoteDisconnected: Remote end closed connection without response
it might be an issue with the census api

code ran the second time I ran it.

code space appears to work. All of the errors that I found were replicated outside of codespace