I am trying to make a subselection of a dataframe based on some columns, while at the same time filtering the dataframe based on a different column. In SQL it looks like this:
Use DataFrame.loc with combination with boolean indexing:
df.loc[df.colume_4 == some_value, ['col1','col2','col3']]
An alternative way:
df[df['col4']==3].filter(['col1', 'col2', 'col3'])