Hello everyone,

This week (5 July - 11 July), This week, I updated the github-actions workflow which was failing earlier, and then created a new workflow, created some functions which would read and download the datasets present in the tidycensus package and had my weekly standup meetings.

 

Meeting 1 (5th July)

During the meeting, I presented the changes I made in the Dockerfile, which resolved the issue, where the tests were failing for every PR. It was resolved by adding a line apt-get install -y --force-yes libpq-dev, as this dependency was required for installing the psycopg2-binary library. As this issue was resolved, I tried making a new workflow in github-actions, which would not require the Dockerfile and the docker-compose.yml.

After the tests were passing, all pending pull requests were merged into the main branch. Merged pull requests which had my contribution are:

 

Meeting 2 (8th July)

During the meeting, I presented my work of how we can access the datasets present in the tidycensus package. Using rpy2, we can run R code in python. So, first task was to create a function which would install the library tidycensus and its dependent libraries. Since some dependent libraries require some additional system dependencies according to the OS, we would prompt the user for such errors.

After that I created a function which would install the tabular datasets fips_codes, pums_variables and mig_recodes. In this function we would call the converter in rpy2 to convert R dataframes into Pandas dataframes. But there was one issue with this function, the mig_recodes dataset caused an encoding error in rpy2’s converter, which reads only utf-8 encoded text. The mig_recodes dataset was encoded with the Windows cp-1252 encoding.

I created a new workflow and created a pull request for it. The PR#1603 Updated python-package.yml is merged.

See you guys next week!!