"The Earth is what we all have in common."

-Wendell Berry

Disclamer: I know some pieces of code are messy yet, for now my purpose is to test approaches, then to clean up the code

Humanity develops civilizations quickly, discover new technologies rapidly, but relatively recently we stopped and started to think what we do to our Earth. Air pollutions, oceans depletions, erroded soils is just a few terms humans are accounted for. Days when we will happily live on Mars do not seem close (and who knows how we might disrupt this planet too) so the best way is to care about our Earth.

My purpose with this project is to explore trends of air pollution, not to tell what we should do. But if at least someone will stop and think about our impact on the world around my work will not be in wain.

I was switching back and forth between data, the main reason for it being my insentive. I wanted a model that would predict pollutant gases for particular region but was not able to find any good dataset for it. Finally, EPA provides tones of data for USA. Here I will use a subset of the data. Full database can be found at https://aqs.epa.gov/aqsweb/airdata/download_files.html. Lookup for county codes you can at https://aqs.epa.gov/aqsweb/documents/codetables/states_and_counties.html Precisely the one I am using I uploaded to Kaggle and it is attached to the notebook.