search

ijyliu / ECMA-31330-Project

Econometrics and Machine Learning Group Project
2 stars 1 forks source link

Visualizing/Presenting Sim Results #24

Closed ijyliu closed 3 years ago

ijyliu commented 3 years ago

There are a ton a of parameters we vary, more than, for example the first problem set where it was easy to set up a seaborn plot grid

ijyliu commented 3 years ago

For the key tables we could definitely do a seaborn plot grid instead of just listing coefficients and standard deviations. Basically we could just have the top two rows of each table in the sim section.

This would also allow us to show the actual distribution and prevent finnicky usage of LaTeX tables that have to be manually updated whenever we change simulation parameters. And I don't think there would be any loss of information. We could easily put a line for the location of the true coefficient.

ijyliu commented 3 years ago

No need to do this, especially since Bonhomme liked the tables