Extracting specific selected columns to new DataFrame as a copy

Question

I have a pandas DataFrame with 4 columns and I want to create a new DataFrame that only has three of the columns. This question is similar to: Extracting specific columns from a data frame but for pandas not R. The following code does not work, raises an error, and is certainly not the pandas way to do it.

import pandas as pd
old = pd.DataFrame({'A' : [4,5], 'B' : [10,20], 'C' : [100,50], 'D' : [-30,-50]})
new = pd.DataFrame(zip(old.A, old.C, old.D)) 
# raises TypeError: data argument can't be an iterator 

What is the pandas way to do it?

Wellchosen

I would like to see a solution that works also if two columns are the same. For example: A, B, B.. – Mulcahy

Answer 1

There is a way of doing this and it actually looks similar to R

new = old[['A', 'C', 'D']].copy()

Here you are just selecting the columns you want from the original data frame and creating a variable for those. If you want to modify the new dataframe at all you'll probably want to use .copy() to avoid a SettingWithCopyWarning.

An alternative method is to use filter which will create a copy by default:

new = old.filter(['A','B','D'], axis=1)

Finally, depending on the number of columns in your original dataframe, it might be more succinct to express this using a drop (this will also create a copy by default):

new = old.drop('B', axis=1)

Cullan

A caution if just copying one column: In old[['A']].copy(), the double square brackets are required to create a new data frame. Note that old['A'].copy() will only create a Series. – Inseverable

Thank you for this amazing explanation. I was having a lot of trouble with creating new columns (after creating a new dataframe without the copy method). Stumped me. You answer finally helped me get to the bottom of it. – Corwin

Answer 2

The easiest way is

new = old[['A','C','D']]

.

Vice

This isn't making a copy unless you explicitly call .copy() – Hartle

this copies by default. – Negation

@Nguaial the behaviour of simple indexing is not specified. You will not know if you get a copy or a view. See documentation for more details: pandas.pydata.org/pandas-docs/stable/user_guide/… – Buckwheat

As mentioned in the comment above, this will create a view and not a copy. – Graybeard

Answer 3

Another simpler way seems to be:

new = pd.DataFrame([old.A, old.B, old.C]).transpose()

where old.column_name will give you a series. Make a list of all the column-series you want to retain and pass it to the DataFrame constructor. We need to do a transpose to adjust the shape.

In [14]:pd.DataFrame([old.A, old.B, old.C]).transpose()
Out[14]: 
   A   B    C
0  4  10  100
1  5  20   50

Inhaler

works, but not if column_name has special characters. – Wendel

oh had not thought of that – Inhaler

@Wendel in that case you can do old['column_name'] I believe – Checkers

@Checkers yes, but that is not in the solution – Wendel

Answer 4

columns by index:

# selected column index: 1, 6, 7
new = old.iloc[: , [1, 6, 7]].copy() 

Charlatan

Answer 5

As an alternative:

new = pd.DataFrame().assign(A=old['A'], C=old['C'], D=old['D'])

Abscess

Answer 6

As far as I can tell, you don't necessarily need to specify the axis when using the filter function.

new = old.filter(['A','B','D'])

returns the same dataframe as

new = old.filter(['A','B','D'], axis=1)

Viosterol

Answer 7

Generic functional form

def select_columns(data_frame, column_names):
    new_frame = data_frame.loc[:, column_names]
    return new_frame

Specific for your problem above

selected_columns = ['A', 'C', 'D']
new = select_columns(old, selected_columns)

Perception

Answer 8

If you want to have a new data frame then:

import pandas as pd
old = pd.DataFrame({'A' : [4,5], 'B' : [10,20], 'C' : [100,50], 'D' : [-30,-50]})
new=  old[['A', 'C', 'D']]

Yep

Dangerous; this isn't making a copy. – Groggy

What's the diffrence between copy and copy of a slice of Dataframe? – Zymometer

Answer 9

You can drop columns in the index:

df = pd.DataFrame({'A': [1, 1], 'B': [2, 2], 'C': [3, 3], 'D': [4, 4]})

df[df.columns.drop(['B', 'C'])]

or

df.loc[:, df.columns.drop(['B', 'C'])]

Output:

   A  D
0  1  4
1  1  4

Chipmunk

Answer 10

You can also use get() to create a new copy (that doesn't run into SettingWithCopyWarning).

new = old.get(['A', 'C', 'D'])

Also, filter selects by column labels by default, so the following works.

new = old.filter(['A', 'C', 'D'])

axis= is needed if one needs to select by row. For example, old.filter([0], axis=0) selects the first row.

If new is an already existing dataframe, then assign() also works (if you want to keep the old columns with their original column names).

new = pd.DataFrame()
new = new.assign(**old[['A', 'C', 'D']])

Diamine