EDA of AQI India Dataset with Tableau and Python

Keynotes:

1. Preprocessed the data in Python

   • Outliers were treated by throttling to max value allowed for each variable.
   • Missing values were replaced by median in case of numerical variables and mode in case of categorical variables.

2. Loaded preprocessed data in Tableau Desktop

3. Changed default aggregation of measures from sum to average because they are measured in average over a period of time.

4. Chose 12 megacities out of 26 which are spread across the country for getting an overall idea of the AQI in India.

5. Bar charts for every pollutant in those top 12 megacities in pre and post covid times.

6. Found decrease in AQI and other pollutants after covid hit, which is good for the environment.