How to sort pandas data frame using values from several columns?

I have the following data frame:

df = pandas.DataFrame([{'c1':3,'c2':10},{'c1':2, 'c2':30},{'c1':1,'c2':20},{'c1':2,'c2':15},{'c1':2,'c2':100}])

Or, in human readable form:

   c1   c2
0   3   10
1   2   30
2   1   20
3   2   15
4   2  100

The following sorting-command works as expected:

df.sort(['c1','c2'], ascending=False)

Output:

   c1   c2
0   3   10
4   2  100
1   2   30
3   2   15
2   1   20

But the following command:

df.sort(['c1','c2'], ascending=[False,True])

results in

   c1   c2
2   1   20
3   2   15
1   2   30
4   2  100
0   3   10

and this is not what I expect. I expect to have the values in the first column ordered from largest to smallest, and if there are identical values in the first column, order by the ascending values from the second column.

Does anybody know why it does not work as expected?

ADDED

This is copy-paste:

>>> df.sort(['c1','c2'], ascending=[False,True])
   c1   c2
2   1   20
3   2   15
1   2   30
4   2  100
0   3   10

DataFrame.sort is deprecated; use DataFrame.sort_values.

>>> df.sort_values(['c1','c2'], ascending=[False,True])
   c1   c2
0   3   10
3   2   15
1   2   30
4   2  100
2   1   20
>>> df.sort(['c1','c2'], ascending=[False,True])
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/ampawake/anaconda/envs/pseudo/lib/python2.7/site-packages/pandas/core/generic.py", line 3614, in __getattr__
    return object.__getattribute__(self, name)
AttributeError: 'DataFrame' object has no attribute 'sort'

Use of sort can result in warning message. See github discussion. So you might wanna use sort_values, docs here

Then your code can look like this:

df = df.sort_values(by=['c1','c2'], ascending=[False,True])

The dataframe.sort() method is - so my understanding - deprecated in pandas > 0.18. In order to solve your problem you should use dataframe.sort_values() instead:

f.sort_values(by=["c1","c2"], ascending=[False, True])

The output looks like this:

    c1 	c2
 	3 	10
 	2 	15
 	2 	30
 	2 	100
 	1 	20

In my case, the accepted answer didn't work:

f.sort_values(by=["c1","c2"], ascending=[False, True])

Only the following worked as expected:

f = f.sort_values(by=["c1","c2"], ascending=[False, True])

If you are writing this code as a script file then you will have to write it like this:

df = df.sort(['c1','c2'], ascending=[False,True])

I have found this to be really useful:

df = pd.DataFrame({'A' : range(0,10) * 2, 'B' : np.random.randint(20,30,20)})

# A ascending, B descending
df.sort(**skw(columns=['A','-B']))

# A descending, B ascending
df.sort(**skw(columns=['-A','+B']))

Note that unlike the standard columns=,ascending= arguments, here column names and their sort order are in the same place. As a result your code gets a lot easier to read and maintain.

Note the actual call to .sort is unchanged, skw (sortkwargs) is just a small helper function that parses the columns and returns the usual columns= and ascending= parameters for you. Pass it any other sort kwargs as you usually would. Copy/paste the following code into e.g. your local utils.py then forget about it and just use it as above.

# utils.py (or anywhere else convenient to import)
def skw(columns=None, **kwargs):
    """ get sort kwargs by parsing sort order given in column name """
    # set default order as ascending (+)
    sort_cols = ['+' + col if col[0] != '-' else col for col in columns]
    # get sort kwargs
    columns, ascending = zip(*[(col.replace('+', '').replace('-', ''), 
                                False if col[0] == '-' else True) 
                               for col in sort_cols])
    kwargs.update(dict(columns=list(columns), ascending=ascending))
    return kwargs

Note : Everything up here is correct,just replace sort --> sort_values() So, it becomes:

 import pandas as pd
 df = pd.read_csv('data.csv')
 df.sort_values(ascending=False,inplace=True)

Refer to the official website here.