Map column from different dataframes with Pandas

There is a time when 2 dataframes need to be summarized in a single table.

For example, a fact table needs to get extra descriptions from outer sources.

Using these tables as an example:

fact_sales_daily

scraped_items

Say that you have a fact table that needs extra description. This extra table does not exist in the database. Probably a co-worker gives it to you in an excel file.

A very naive way is to loop and assign by row. But that will take a long time if the table is huge.

You can try this way.

Step 1: Create dict of item_id and value you want to map

item_source_dict = dict(zip(scraped_items['item_id'], scraped_items['source']))

Step 2: Map the column to the dict

 fact_sales_daily['item_source'] = fact_sales_daily['item_id'].map(item_source_dict)

This step maps the value of the item to its value from the dict. And the result is stored in a column item_source.

You will get something like this

Simple and readable.