Pandas reindex subset of rows-CodePudding

I have the following dataframe:

import pandas as pd
#Create DF
d = { 
     'Country': ['USA','USA','AUS','AUS','AUS','UK','UK'],
    'poulation_k':[200,250,150,120,350,800,600,],
    }
           
df = pd.DataFrame(data=d)
df

I would like to sort the rows by poulation where Country = AUS but maintaining their order in the overall dataframe:

So my expected output will be:

I would also like to do it by the other countries however i would like to do it on a manual basis - i.e i would like the function to specify the Country name. Any help would be fantastic! Thanks

UPDATE:

Edit, i negleted to put in that my dataframes consist of other columns as well. so the example should be:

import pandas as pd
#Create DF
d = { 
     'Country': ['USA','USA','AUS','AUS','AUS','UK','UK'],
    'poulation_k':[200,250,150,120,350,800,600,],
    }
           
df = pd.DataFrame(data=d)
df['Test'] = df.poulation_k.astype(str)   df.Country
df

With the expected output of the function to sort AUS being:

CodePudding user response：

def func(country):
    sub_df = df[df['Country']==country] 

    idx_i = sub_df.index
    idx_f = sub_df.sort_values('poulation_k', ascending=False).index

    return df.rename(dict(zip(idx_f, idx_i))).sort_index()
    
func('AUS')

My solution modifies the index so it will work no matter how many columns you have in your dataframe.

CodePudding user response：

As you only reorder one column, you can do it simply:

df.loc[df['Country'] == 'AUS', 'poulation_k'] = np.sort(
    df.loc[df['Country'] == 'AUS', 'poulation_k'].values)

It gives as expected:

  Country  poulation_k
0     USA          200
1     USA          250
2     AUS          120
3     AUS          150
4     AUS          350
5      UK          800
6      UK          600