Reindexing in Pandas can be used to change the index of rows and columns of a DataFrame. Indexes can be used with reference to many index DataStructure associated with several pandas series or pandas DataFrame. Let’s see how can we Reindex the columns and rows in Pandas DataFrame.
Reindexing the Rows
One can reindex a single row or multiple rows by using reindex() method. Default values in the new index that are not present in the dataframe are assigned NaN.
Example #1:
Python3
# import numpy and pandas moduleimport pandas as pdimport numpy as npcolumn=['a','b','c','d','e']index=['A','B','C','D','E']# create a dataframe of random values of arraydf1 = pd.DataFrame(np.random.rand(5,5), columns=column, index=index)print(df1)print('\n\nDataframe after reindexing rows: \n',df1.reindex(['B', 'D', 'A', 'C', 'E'])) |
Output:
a b c d e
A 0.129087 0.445892 0.898532 0.892862 0.760018
B 0.635785 0.380769 0.757578 0.158638 0.568341
C 0.713786 0.069223 0.011263 0.166751 0.960632
D 0.913553 0.676715 0.141932 0.202201 0.346274
E 0.050204 0.132140 0.371349 0.633203 0.791738
Dataframe after reindexing rows:
a b c d e
B 0.635785 0.380769 0.757578 0.158638 0.568341
D 0.913553 0.676715 0.141932 0.202201 0.346274
A 0.129087 0.445892 0.898532 0.892862 0.760018
C 0.713786 0.069223 0.011263 0.166751 0.960632
E 0.050204 0.132140 0.371349 0.633203 0.791738
Example #2:
Python3
# import numpy and pandas moduleimport pandas as pdimport numpy as npcolumn = ['a', 'b', 'c', 'd', 'e']index = ['A', 'B', 'C', 'D', 'E'] # create a dataframe of random values of arraydf1 = pd.DataFrame(np.random.rand(5, 5), columns = column, index = index)# create the new index for rowsnew_index =['U', 'A', 'B', 'C', 'Z']print(df1.reindex(new_index)) |
Output:
a b c d e
U NaN NaN NaN NaN NaN
A 0.523572 0.378545 0.871649 0.980319 0.397569
B 0.796003 0.516602 0.839177 0.835811 0.831672
C 0.160613 0.833154 0.810910 0.771017 0.225579
Z NaN NaN NaN NaN NaN
Reindexing the columns using the axis keyword
One can reindex a single column or multiple columns by using reindex() method and by specifying the axis we want to reindex. Default values in the new index that are not present in the dataframe are assigned NaN.
Example #1:
Python3
# import numpy and pandas moduleimport pandas as pdimport numpy as npcolumn=['a','b','c','d','e']index=['A','B','C','D','E']#create a dataframe of random values of arraydf1 = pd.DataFrame(np.random.rand(5,5), columns=column, index=index)column=['e','a','b','c','d'] # create the new index for columnsprint(df1.reindex(column, axis='columns')) |
Output:
e a b c d
A 0.592727 0.337282 0.686650 0.916076 0.094920
B 0.235794 0.030831 0.286443 0.705674 0.701629
C 0.882894 0.299608 0.476976 0.137256 0.306690
D 0.758996 0.711712 0.961684 0.235051 0.315928
E 0.911693 0.436031 0.822632 0.477767 0.778608
Example #2:
Python3
# import numpy and pandas moduleimport pandas as pdimport numpy as npcolumn =['a', 'b', 'c', 'd', 'e']index =['A', 'B', 'C', 'D', 'E'] # create a dataframe of random values of arraydf1 = pd.DataFrame(np.random.rand(5, 5), columns = column, index = index)column =['a', 'b', 'c', 'g', 'h']# create the new index for columnsprint(df1.reindex(column, axis ='columns')) |
Output:
a b c g h
A 0.390460 0.795073 0.369077 NaN NaN
B 0.855556 0.856980 0.132092 NaN NaN
C 0.662565 0.230554 0.215567 NaN NaN
D 0.712128 0.424346 0.813452 NaN NaN
E 0.543142 0.847750 0.168018 NaN NaN
Replacing the missing values
Code #1: Missing values from the dataframe can be filled by passing a value to the keyword fill_value. This keyword replaces the NaN values.
Python3
# import numpy and pandas moduleimport pandas as pdimport numpy as npcolumn =['a', 'b', 'c', 'd', 'e']index =['A', 'B', 'C', 'D', 'E'] # create a dataframe of random values of arraydf1 = pd.DataFrame(np.random.rand(5, 5), columns = column, index = index)column =['a', 'b', 'c', 'g', 'h']# create the new index for columnsprint(df1.reindex(column, axis ='columns', fill_value = 1.5)) |
Output:
a b c g h
A 0.945594 0.492603 0.705738 1.5 1.5
B 0.794345 0.068308 0.017898 1.5 1.5
C 0.622142 0.880565 0.035528 1.5 1.5
D 0.577288 0.934063 0.824655 1.5 1.5
E 0.636026 0.316232 0.244597 1.5 1.5
Code #2: Replacing the missing data with a string.
Python3
# import numpy and pandas moduleimport pandas as pdimport numpy as npcolumn =['a', 'b', 'c', 'd', 'e']index =['A', 'B', 'C', 'D', 'E'] # create a dataframe of random values of arraydf1 = pd.DataFrame(np.random.rand(5, 5), columns = column, index = index)column =['a', 'b', 'c', 'g', 'h']# create the new index for columnsprint(df1.reindex(column, axis ='columns', fill_value ='data missing')) |
Output:
a b c g h
A 0.227380 0.809179 0.879175 data missing data missing
B 0.212493 0.335610 0.306006 data missing data missing
C 0.406346 0.852985 0.422182 data missing data missing
D 0.145821 0.648285 0.004842 data missing data missing
E 0.002305 0.694541 0.657602 data missing data missing
