Closed juzcho closed 3 years ago
The stats models library seems like it has a lot of potential to me. here is the link with the write up and functions, not all of them would be applicable seeing as we arent doing something time series related but some of them look pretty promising to me
This is related to #9, so closing #9 in favor of this.
If you guys want to use the small kickstarter datasets. I will need to use the fixer.io API to convert the money cols to USD based to match our other dfs. So I will import and use a fixer.io package with:
pip install fixerio
easy_install fixerio
from fixerio import Fixerio
Link for reference:
https://fixerio.readthedocs.io/en/latest/ https://fixer.io/documentation
Papermill to seperate code: https://papermill.readthedocs.io/en/latest/usage-execute.html
Ended up not using fixer.io. It does not really make sense to use on the small kickstarter data because we don't have any launched or deadline dates.
Adding viz libs to close this issue. Here is a link of what I added. https://towardsdatascience.com/top-6-python-libraries-for-visualization-which-one-to-use-fe43381cd658
Each graphs different things better. Look at the related sites to see which one you want to use. We can remove whatever we don't use later.
For each do the following to your dev env. @cdhendy @juzcho We should add the below to the readme later.
Seaborn: conda install seaborn
Altair: conda install -c conda-forge altair
Bokeh: conda install bokeh
Folium: conda install folium -c conda-forge
Major Requirement for the Project.