Pandas Split DataFrame using row index

Question

I want to split dataframe by uneven number of rows using row index.

The below code:

groups = df.groupby((np.arange(len(df.index))/l[1]).astype(int))

works only for uniform number of rows.

df

a b c  
1 1 1  
2 2 2  
3 3 3  
4 4 4  
5 5 5  
6 6 6  
7 7 7  

l = [2, 5, 7]

df1  
1 1 1  
2 2 2  

df2  
3,3,3  
4,4,4  
5,5,5  

df3  
6,6,6  
7,7,7  

df4  
8,8,8

Answer 1

You could use list comprehension with a little modications your list, l, first.

print(df)

   a  b  c
0  1  1  1
1  2  2  2
2  3  3  3
3  4  4  4
4  5  5  5
5  6  6  6
6  7  7  7
7  8  8  8


l = [2,5,7]
l_mod = [0] + l + [max(l)+1]

list_of_dfs = [df.iloc[l_mod[n]:l_mod[n+1]] for n in range(len(l_mod)-1)]

Output:

list_of_dfs[0]

   a  b  c
0  1  1  1
1  2  2  2

list_of_dfs[1]

   a  b  c
2  3  3  3
3  4  4  4
4  5  5  5

list_of_dfs[2]

   a  b  c
5  6  6  6
6  7  7  7

list_of_dfs[3]

   a  b  c
7  8  8  8

Answer 2

I think this is what you need:

df = pd.DataFrame({'a': np.arange(1, 8),
                  'b': np.arange(1, 8),
                  'c': np.arange(1, 8)})
df.head()
    a   b   c
0   1   1   1
1   2   2   2
2   3   3   3
3   4   4   4
4   5   5   5
5   6   6   6
6   7   7   7

last_check = 0
dfs = []
for ind in [2, 5, 7]:
    dfs.append(df.loc[last_check:ind-1])
    last_check = ind

Although list comprehension are much more efficient than a for loop, the last_check is necessary if you don't have a pattern in your list of indices.

dfs[0]

    a   b   c
0   1   1   1
1   2   2   2

dfs[2]

    a   b   c
5   6   6   6
6   7   7   7