search

jforbes14 / eechidna

Data package containing tables, shapefiles, etc. for the 2001-2019 Australian elections and Censuses
https://jforbes14.github.io/eechidna/
43 stars 6 forks source link

2019 data #30

Closed danmackinlay closed 3 years ago

danmackinlay commented 3 years ago

Over the summer holidays 'd like to pull in the 2019 election results to the project. Before I do so, is there anything I should know about best practices for this? The format should, I suppose be much like the 2016 results.

danmackinlay commented 3 years ago

Perhaps @emitanaka specifically could let me know: are you or your students already doing this? I notice your fork has some updates

emitanaka commented 3 years ago

@danmackinlay I'm not working actively on it so feel free to ignore mine! I just needed a fix to the vic map since the original seemed to have issues but I don't have much time to do a proper test for my code addition, so I don't plan on pushing my changes any time soon.

danmackinlay commented 3 years ago

It looks like you did solid work though, will try to see if I have enough time to take it anywhere. We need another R unconf to get this done, eh?

jforbes14 commented 3 years ago

@danmackinlay @emitanaka what specifically was the problem here? I'll try to resolve if I can get my head around it

bryceroney commented 3 years ago

Isn't the 2019 data already available in fp19 and tcp19? Either way, happy to help where I can on the project.

danmackinlay commented 3 years ago

Yes the 2019 data is here! It is absent from several places in the documentation, but is in fact there, I discovered upon installation. Possible an imputed version of the ABS data might be wise also, but that is a separate question.

I did, however, notice that there is a glitch in the 2019 data. Specifically, the Greens party is missing the Greens Victoria, so I had to fix that before I could use the data. Here is what I mean: https://github.com/danmackinlay/eechidna/commit/0deeac67da27be7eeaaf08a2922282f9428ad59a

danmackinlay commented 3 years ago

Cripes I just noticed what a messy diff that is. Looks like it picked up some whitespace normalization when I did the rebase. Sorry! Anyway, I can prepare a clean patch which preserves all the trailing whitespace if desired; the real change is only 2 lines.

jforbes14 commented 3 years ago

Yes a patch would be great @danmackinlay! I'll also make some moves on the ABS imputation for 2019 👍

bryceroney commented 3 years ago

I'm happy to look over updating the documentation to make sure it's referenced correctly.

danmackinlay commented 3 years ago

Possibly a side issue, or a separate documentation ticket, but what is the canonical way to regenerate the package data after changing the import scripts? If I do what I assume the right thing is and source the script data-raw/all-electorate-votes.R from the package root it seems to do the right thing but it also throws some errors about missing names in the ABS tables. Anyway, by that method I updated some data - see https://github.com/danmackinlay/eechidna/commit/624a58d0c3f650911861b7c6f742ea87a1b56eae

but I would like to know I'm not breaking anything before I make a new pull request with other similar updates.

jforbes14 commented 3 years ago

@danmackinlay I've merged your PR and have updated the 2019 election data on master. I'm working on a new release for this, so the CRAN version will also be updated soon (also pending documentation review @bryceroney ).

Some of the wrangling scripts including data-raw/all-electorate-votes.R are a bit confusing because they are expecting objects mid-wrangle. Probably an exercise for me to go and do some refactoring.

danmackinlay commented 3 years ago

Should we close this issue in favour of openeing new tickets for new issues as they arise?