Project begins by using selenium (python api testing/web-scraping module) to scrape relevant polling data information on every state and project into a text files where the information is organized (ie dates formatted etc)
Using mongodb python API "Pymongo" all infomation that is suitable is added to a database without duplicates. Read Operations are performed to create average Trump/Biden polling data values for all states that have information available. Thes figures are then projected as a dictionary and added into the dbf file which is edited using geopandas/pandas and pythons csv module.
Using matplotlib, geopandas and numpy the mapping data is combined with the election data gathered to create a color coded map based on aggregated percentage polling values.
The map has a series of colors
- The light green indicates states that do not have sufficient polling data to be projected.
- Anything blue are states leaning democratic
- Anything red are states leaning republican
- Anything grey is a tie.
- This map is of polling data from June 20th
All of these polling values are subject to change over time.
I found the initial map data at this site.. Initial Election data
This is what the map looks like initailly projected
I then used various algorithms and projection methods that are found in Arc Gis software to change this map to better suit my needs
Color coding and resizing of the projection was completed with matplotlib.
List of all resources used
- Selenium
- Matplotlib
- Numpy
- GeoPandas, Pandas, csv edit etc
- Arc Gis
- Mongodb/Pymongo