iterate through pandas dataframe consecutive columns

Question

I am trying to create a loop in pandas to calculate difference between consecutive columns and give output in a new column:

Original df:

**201601** **201602** **201603**  
100           200         500

Desired output

**201601** **201602** **201603**  **201602_201601** **201603_02**
100           200         500         100          300

My code is which I had modified from a stackoverflow post ([ add columns to a data frame calculated by for loops in python ):

for i in df.iloc[:,2:5]:
  for j in df.iloc[:,2:5]:
    if i == j:
        break
    else:
        bina = df[i]-df[j]
        df['MOM_' + str(j) + '_' + str(i)] = bina
df.head()

However, the output I get is as below:

**201601** **201602** **201603**  **201602_201601** **201603_201601** **201603_201602**
100           200         500         100          400   300

I have used pd.diff to do what I needed but couldn't figure out the for loop code. Any help would be greatly appreciated.

Thanks

Answer 1

Using diff and simple list comprehension with zip to construct the columns' names.

cols = [f'{b}_{a}' for (a,b) in zip(df.columns, df.columns[1:])]
df[cols] = df.diff(axis=1).dropna(axis=1)

    201601  201602  201603  201602_201601   201603_201602
0   100     200     500     100             300

Avoid to use for loops at all times when using pandas

Answer 2

This is just fix your code

col=df.columns
for x,i in enumerate(col):
    for y,j in enumerate(col):
        if  y-x==1 and i!=j:
            bina = df[i]-df[j]
            df['MOM_' + str(j) + '_' + str(i)] = bina
df.columns
Out[1210]: 
Index(['**201601**', '**201602**', '**201603**', 'MOM_**201602**_**201601**',
       'MOM_**201603**_**201602**'],
      dtype='object')