Cheat sheets/notes on EDA step by step

Question

I’m facing difficulties in analysing the data when to use what commands/how to proceed to next step.

For example :-)

1. I don’t know when to remove which column/row data

2. when to use scatter, histo, box plots

3. when remove duplicate values, when to add median, mean for which data like uni or bi

4. After plotting graphs when to use heat maps, when to use subplots


Please do post some resources which can help to understand step by step process when to use what in analysing different datasets

Answer 1

This is a cool thought provoking post but EDA is where the most individuality exists in data science. It's just you exploring.

I doubt many will have a comprehensive answer for you

Answer 2

There are no golden rules for EDA. Indeed, we know when to use scatter plot and box plots. However, the key is what messages you are trying to convey. How you are building the visualizations should be highly dependent on this.

Answer 3

run a pandas profiling  report