Sum of different slices rows and column

Question

I have pandas DataFrame df and three arrays columns_list, lower_boarder and upper_boarder all have the same shape. I want to find array with shape as input arrays. For i-th sum in array I need sum of i-th column from i-th lower_boarder and i-th upper_boarder. Is it possible to do in pandas way?

df = pd.DataFrame(np.arange(0, 10).repeat(5).reshape(10, 5), columns=['col1', 'col2', 'col3', 'col4', 'col5'])


      col1  col2  col3  col4  col5
0     0     0     0     0     0
1     1     1     1     1     1
2     2     2     2     2     2
3     3     3     3     3     3
4     4     4     4     4     4
5     5     5     5     5     5
6     6     6     6     6     6
7     7     7     7     7     7
8     8     8     8     8     8
9     9     9     9     9     9

columns_list = ['col1', 'col2', 'col3', 'col1']
lower_boarder = [2, 3, 5, 6]
upper_boarder = [4, 5, 6, 9]

sum = []
for i in range(len(columns_list)):
    sum.append(df.loc[lower_boarder[i]:upper_boarder[i], columns_list[i]].sum())

How can I do it without itteration? If shapes of columns_list is different?

Answer 1

Create DataFrame only by columns by columns_list by DataFrame.reindex and then compare arange by length of DataFrame by lower_boarder, upper_boarder with numpy broadcasting, if not match generate 0 by DataFrame.where, last sum columns to Series:

df1 = df.reindex(columns_list, axis=1)
m1 = np.arange(len(df1))[:, None] >= np.array(lower_boarder) 
m2 = np.arange(len(df1))[:, None] <= np.array(upper_boarder) 

#if default index
m1 = df1.index.to_numpy()[:, None] >= np.array(lower_boarder) 
m2 = df1.index.to_numpy()[:, None] <= np.array(upper_boarder)

---

s = df1.where(m1 & m2, 0).sum()
print (s)
col1     9
col2    12
col3    11
col1    30
dtype: int64

Numpy alternative with numpy.where:

a = np.where(m1 & m2, df1, 0).sum(axis=0).tolist()
print (a)
[9, 12, 11, 30]