Calculating fold change for each element between data frames-CodePudding

I have 2 data frames of equal number of columns and rows (NxM). I'm looking to calculate fold change element-wise. So as to each element in data frame 2 gets subtracted with the corresponding element in data frame 1 and divided by the corresponding element in data frame 1. I'm leaving 2 example data frames below with only 2 columns but my data frames have 150 columns and 1000 rows. I'm having trouble writing code for it.

Col 1.	Col 2.
1.	3.
4.	5.
7.	1.
1.	9.

Col 1.	Col 2.
1.	3.
4.	5.
1.	5.
7.	1.

This is what ive been doing to subtract them:

#bootstrapping all columns
df_sub1 = data1.sample( axis='rows',  n=1000, replace = True, ignore_index = True)
df_sub7 = data7.sample( axis='rows',  n=1000, replace = True, ignore_index = True)
#subtracting data frames element-wise
df_fc = df_sub7.sub(df_sub1)

How could I go about turning this snippet into fold change for each element? Thank you!

CodePudding user response：

element wise division between two dataframes can be expressed as df1 / df2. note however, that pandas will attempt to align the indexes. this is not an issue if the dataframes are identically indexed.

using the sample data provided.

df1 = pd.DataFrame({
  'c1': [1,4,7,1],
  'c2': [3,5,1,9]
  })

df2 = pd.DataFrame({
  'c1': [1,4,1,7],
  'c2': [3,5,5,1]
  })

folded = (df2 - df1) / df1
# alternatively:
folded = df2.sub(df1).div(df1)

folded has the following values

         c1        c2
0  0.000000  0.000000
1  0.000000  0.000000
2 -0.857143  4.000000
3  6.000000 -0.888889