I need to perform a regression analysis in python with a rather difficult dataset. I have 13 entities with each 48 variables for 17 observation periods. In the end, I need one model which can predict a certain variable with the data from those 13 entities and their variables.

Firstly, I want to test which of those variables are overall significant and which are not. Those insignificant should not be included in the model.
After the testing procedure, I need the model and its r2 value.

Does somebody know how to achieve this? I cant resolve it myself and neither with the help of AI..

The dataset includes in the first column the entities/groups, the second colum the observation periods, and the rest of the columns are the variables.

Thank you for helping!