Choosing columns in pandas DataFrame

Sometimes, you have a lot of columns in your DataFrame and want to use only some of them.

Picking specific columns

df[‘col1]

This command picks a column and returns it as a Series

Column returned as a Series

df[[‘col1’]]

Here, I chose the column and I get a DataFrame

Column returned as a DataFrame

df[[‘col1’, ‘col2’]]

This is the same command as above — but this time, I am choosing more than one column

Two columns returned as a DataFrame

Picking certain values from a column

df[df1[‘col1’] == value]

You choose all of the values in column 1 that are equal to the value.

df[df1[‘col1’] != value]

All of the values in column 1 that are not equal to the value.

df[df1[‘col1’] < value]

All of the values in column 1 are smaller than the value.

df[df1[‘col1’] > value]

All of the values in column 1 are bigger than the value.

All of the values in ‘Police District’ are 8

df1[‘col1’] == value

Similarly to the above commands, just here you get Boolean values.

The column returned as boolean

Picking certain rows

df.ix[index]

You can actually choose a row by using its index (the number on the far left).

I chose a row with index ‘0’

df.ix[‘index name’]

This command does exactly the same thing as above but you use it when you have actually named your indices.

Deleting columns

df.drop([‘col1’,’col2'], axis=1)

This command deletes specific columns from the DataFrame

You can see that here I don’t have the two rows that I dropped

del df[‘col1’]

This command deletes a specific column from the DataFrame and modifies it — so be careful how you use it.

You can see that this column is now gone

All of the code can be found on my GitHub: https://github.com/kasiarachuta/Blog/blob/master/Choosing%20specific%20columns%20or%20rows%20from%20pandas%20DataFrame.ipynb

from : https://medium.com/@kasiarachuta/choosing-columns-in-pandas-dataframe-d0677b34a6ca