How to compare two dataframes element-wise?-CodePudding

I want to compare two dataframes in an element-wise manner in order to find which elements in dataframe one are less than those in dataframe two.

DataFrame One:

DataFrame Two:

In this case the only elements that should return True should be elements on Column 130_x for Indexes 0 and 1. The rest should return False.

Any help would be appreciated.

CodePudding user response：

Then you need to change the column names

df1.columns = df2.columns

Or here we use replace

df1.columns = df1.columns.str.replace('_x','_y')
s = df1.eq(df2).all(1)
s.index[s]

CodePudding user response：

You could simply use df1 < df2 but your column names needs to be the same.

import pandas as pd
from io import StringIO

data1 = StringIO("""
126_x,127_x,128_x,129_x,130_x
0.0,0.04898,0.359184,0.004082,0.134694
0.0,0.04898,0.359184,0.004082,0.134694
""")

data2 = StringIO("""
126_y,127_y,128_y,129_y,130_y
0.0,0.0,0.175439,0.0,0.166667
0.006579,0.0,0.269737,0.0,0.223684
""")

df1 = pd.read_csv(data1)
df2 = pd.read_csv(data2)

# column header needs to be the same
df2.columns = df1.columns

print(df1 < df2)

#    126_x  127_x  128_x  129_x  130_x
# 0  False  False  False  False   True
# 1   True  False  False  False   True

In this case the only elements that should return True should be elements on Column 130_x for Indexes 0 and 1. The rest should return False.

That's not correct, there is another cell (Column 126_y, Row 2) which would be true:
0.006579 > 0.0.

CodePudding user response：

You don't need to modify the dataframes. You can use the underlying numpy array:

df1.lt(df2.values)

Output:

   126_x  127_x  128_x  129_x  130_x
0  False  False  False  False   True
1   True  False  False  False   True