If I have a dataframe like this how can I decrement 'a' based on 'b'?

df = pd.DataFrame([[100,1], [100,2], [100,3], [100,4], [100,5]], columns=["a", "b"])
df.loc[df.b > 2, 'b']  = 1
df.loc[df.b % 2 == 0, 'a'] -= 10
df.loc[df.b % 2 != 0, 'a'] = np.NaN

Desired result:

   a    b
0  NaN  1
1   90  2
2   80  4
3  NaN  5
4   70  6

CodePudding user response：

You can subtract values in array created by np.arange starting by 1 multiplied by 10:

df.loc[df.b > 2, 'b']  = 1

m = df.b % 2 == 0
df.loc[m, 'a'] -= np.arange(1, m.sum()   1) * 10
#alternative
#df.loc[m, 'a'] -= np.arange(10, m.sum() * 10   10, 10)
df.loc[~m, 'a'] = np.NaN
print (df)
      a  b
0   NaN  1
1  90.0  2
2  80.0  4
3   NaN  5
4  70.0  6