GROUP BY column in pandas

Posted by Kevin
Aggregating and grouping rows in SQL is performed using GROUP BY. In pandas, .groupby() is applied to achieve the same results.

Grouping rows by a single column in pandas for aggregating data

In SQL:

SELECT
  column_1,
  COUNT(column_2),
  AVG(column_2)
FROM
  table
GROUP BY
  column_1;

In pandas:

tips
    .groupby('column_1')
    .agg({'column_2': ['size', 'mean']})

Grouping rows by multiple columns in pandas for aggregating data

In SQL:

SELECT
  column_1,
  column_3,
  COUNT(column_2),
  AVG(column_2)
FROM
  table
GROUP BY
  column_1,
  column_3;

In pandas:

tips
    .groupby(['column_1', 'column_3'])
    .agg({'column_2': ['size', 'mean']})

References