Efficient way to unnest (explode) multiple list columns in a pandas DataFrame
Question:
I am reading multiple JSON objects into one DataFrame. The problem is that some of the columns are lists. Also, the data is very big and because of that I cannot use the available solutions on the internet. They are very slow and memory-inefficient
Here is how my data looks like:
df = pd.DataFrame({'A': ['x1','x2','x3', 'x4'], 'B':[['v1','v2'],['v3','v4'],['v5','v6'],['v7','v8']], 'C':[['c1','c2'],['c3','c4'],['c5','c6'],['c7','c8']],'D':[['d1','d2'],['d3','d4'],['d5','d6'],['d7','d8']], 'E':[['e1','e2'],['e3','e4'],['e5','e6'],['e7','e8']]})
A B C D E
0 x1 [v1, v2] [c1, c2] [d1, d2] [e1, e2]
1 x2 [v3, v4] [c3, c4] [d3, d4] [e3, e4]
2 x3 [v5, v6] [c5, c6] [d5, d6] [e5, e6]
3 x4 [v7, v8] [c7, c8] [d7, d8] [e7, e8]
And this is the shape of my data: (441079, 12)
My desired output is:
A B C D E
0 x1 v1 c1 d1 e1
0 x1 v2 c2 d2 e2
1 x2 v3 c3 d3 e3
1 x2 v4 c4 d4 e4
.....
EDIT: After being marked as duplicate, I would like to stress on the fact that in this question I was looking for an efficient method of exploding multiple columns. Therefore the approved answer is able to explode an arbitrary number of columns on very large datasets efficiently. Something that the answers to the other question failed to do (and that was the reason I asked this question after testing those solutions).
Answers:
Use set_index on A and on remaining columns apply and stack the values. All of this condensed into a single liner.
In [1253]: (df.set_index('A')
.apply(lambda x: x.apply(pd.Series).stack())
.reset_index()
.drop('level_1', 1))
Out[1253]:
A B C D E
0 x1 v1 c1 d1 e1
1 x1 v2 c2 d2 e2
2 x2 v3 c3 d3 e3
3 x2 v4 c4 d4 e4
4 x3 v5 c5 d5 e5
5 x3 v6 c6 d6 e6
6 x4 v7 c7 d7 e7
7 x4 v8 c8 d8 e8
def explode(df, lst_cols, fill_value=''):
# make sure `lst_cols` is a list
if lst_cols and not isinstance(lst_cols, list):
lst_cols = [lst_cols]
# all columns except `lst_cols`
idx_cols = df.columns.difference(lst_cols)
# calculate lengths of lists
lens = df[lst_cols[0]].str.len()
if (lens > 0).all():
# ALL lists in cells aren't empty
return pd.DataFrame({
col:np.repeat(df[col].values, df[lst_cols[0]].str.len())
for col in idx_cols
}).assign(**{col:np.concatenate(df[col].values) for col in lst_cols})
.loc[:, df.columns]
else:
# at least one list in cells is empty
return pd.DataFrame({
col:np.repeat(df[col].values, df[lst_cols[0]].str.len())
for col in idx_cols
}).assign(**{col:np.concatenate(df[col].values) for col in lst_cols})
.append(df.loc[lens==0, idx_cols]).fillna(fill_value)
.loc[:, df.columns]
Usage:
In [82]: explode(df, lst_cols=list('BCDE'))
Out[82]:
A B C D E
0 x1 v1 c1 d1 e1
1 x1 v2 c2 d2 e2
2 x2 v3 c3 d3 e3
3 x2 v4 c4 d4 e4
4 x3 v5 c5 d5 e5
5 x3 v6 c6 d6 e6
6 x4 v7 c7 d7 e7
7 x4 v8 c8 d8 e8
pandas >= 0.25
Assuming all columns have the same number of lists, you can call Series.explode on each column.
df.set_index(['A']).apply(pd.Series.explode).reset_index()
A B C D E
0 x1 v1 c1 d1 e1
1 x1 v2 c2 d2 e2
2 x2 v3 c3 d3 e3
3 x2 v4 c4 d4 e4
4 x3 v5 c5 d5 e5
5 x3 v6 c6 d6 e6
6 x4 v7 c7 d7 e7
7 x4 v8 c8 d8 e8
The idea is to set as the index all columns that must NOT be exploded first, then reset the index after.
It’s also faster.
%timeit df.set_index(['A']).apply(pd.Series.explode).reset_index()
%%timeit
(df.set_index('A')
.apply(lambda x: x.apply(pd.Series).stack())
.reset_index()
.drop('level_1', 1))
2.22 ms ± 98.6 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
9.14 ms ± 329 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
Building on @cs95’s answer, we can use an if clause in the lambda function, instead of setting all the other columns as the index. This has the following advantages:
- Preserves column order
- Lets you easily specify columns using the set you want to modify,
x.name in [...], or not modify x.name not in [...].
df.apply(lambda x: x.explode() if x.name in ['B', 'C', 'D', 'E'] else x)
A B C D E
0 x1 v1 c1 d1 e1
0 x1 v2 c2 d2 e2
1 x2 v3 c3 d3 e3
1 x2 v4 c4 d4 e4
2 x3 v5 c5 d5 e5
2 x3 v6 c6 d6 e6
3 x4 v7 c7 d7 e7
3 x4 v8 c8 d8 e8
Here is my solution using ‘apply’ function. Main features/differences:
- offer options to specify selected multiple columns or all columns
- offer options to specify values to fill in the ‘missing’ position (through parameter fill_mode = ‘external’; ‘internal’; or ‘trim’, explanation would be long, see examples below and try yourself to change the option and check the result)
Notes: option ‘trim’ was developed for my need, out of scope for this question
def lenx(x):
return len(x) if isinstance(x,(list, tuple, np.ndarray, pd.Series)) else 1
def cell_size_equalize2(row, cols='', fill_mode='internal', fill_value=''):
jcols = [j for j,v in enumerate(row.index) if v in cols]
if len(jcols)<1:
jcols = range(len(row.index))
Ls = [lenx(x) for x in row.values]
if not Ls[:-1]==Ls[1:]:
vals = [v if isinstance(v,list) else [v] for v in row.values]
if fill_mode=='external':
vals = [[e] + [fill_value]*(max(Ls)-1) if (not j in jcols) and (isinstance(row.values[j],list))
else e + [fill_value]*(max(Ls)-lenx(e))
for j,e in enumerate(vals)]
elif fill_mode == 'internal':
vals = [[e]+[e]*(max(Ls)-1) if (not j in jcols) and (isinstance(row.values[j],list))
else e+[e[-1]]*(max(Ls)-lenx(e))
for j,e in enumerate(vals)]
else:
vals = [e[0:min(Ls)] for e in vals]
row = pd.Series(vals,index=row.index.tolist())
return row
Examples:
df=pd.DataFrame({
'a':[[1],2,3],
'b':[[4,5,7],[5,4],4],
'c':[[4,5],5,[6]]
})
print(df)
df1 = df.apply(cell_size_equalize2, cols='', fill_mode='external', fill_value = "OK", axis=1).apply(pd.Series.explode)
print('nfill_mode='external', all columns, fill_value = 'OK'n', df1)
df2 = df.apply(cell_size_equalize2, cols=['a', 'b'], fill_mode='external', fill_value = "OK", axis=1).apply(pd.Series.explode)
print('nfill_mode='external', cols = ['a', 'b'], fill_value = 'OK'n', df2)
df3 = df.apply(cell_size_equalize2, cols=['a', 'b'], fill_mode='internal', axis=1).apply(pd.Series.explode)
print('nfill_mode='internal', cols = ['a', 'b']n', df3)
df4 = df.apply(cell_size_equalize2, cols='', fill_mode='trim', axis=1).apply(pd.Series.explode)
print('nfill_mode='trim', all columnsn', df4)
Output:
a b c
0 [1] [4, 5, 7] [4, 5]
1 2 [5, 4] 5
2 3 4 [6]
fill_mode='external', all columns, fill_value = 'OK'
a b c
0 1 4 4
0 OK 5 5
0 OK 7 OK
1 2 5 5
1 OK 4 OK
2 3 4 6
fill_mode='external', cols = ['a', 'b'], fill_value = 'OK'
a b c
0 1 4 [4, 5]
0 OK 5 OK
0 OK 7 OK
1 2 5 5
1 OK 4 OK
2 3 4 6
fill_mode='internal', cols = ['a', 'b']
a b c
0 1 4 [4, 5]
0 1 5 [4, 5]
0 1 7 [4, 5]
1 2 5 5
1 2 4 5
2 3 4 6
fill_mode='trim', all columns
a b c
0 1 4 4
1 2 5 5
2 3 4 6
As of pandas 1.3.0 (What’s new in 1.3.0 (July 2, 2021)):
DataFrame.explode() now supports exploding multiple columns. Its column argument now also accepts a list of str or tuples for exploding on multiple columns at the same time (GH39240)
So now this operation is as simple as:
df.explode(['B', 'C', 'D', 'E'])
A B C D E
0 x1 v1 c1 d1 e1
0 x1 v2 c2 d2 e2
1 x2 v3 c3 d3 e3
1 x2 v4 c4 d4 e4
2 x3 v5 c5 d5 e5
2 x3 v6 c6 d6 e6
3 x4 v7 c7 d7 e7
3 x4 v8 c8 d8 e8
Or if wanting unique indexing:
df.explode(['B', 'C', 'D', 'E'], ignore_index=True)
A B C D E
0 x1 v1 c1 d1 e1
1 x1 v2 c2 d2 e2
2 x2 v3 c3 d3 e3
3 x2 v4 c4 d4 e4
4 x3 v5 c5 d5 e5
5 x3 v6 c6 d6 e6
6 x4 v7 c7 d7 e7
7 x4 v8 c8 d8 e8
Gathering all of the responses on this and other threads, here is how I do it for comma-delineated rows:
from collections.abc import Sequence
import pandas as pd
import numpy as np
def explode_by_delimiter(
df: pd.DataFrame,
columns: str | Sequence[str],
delimiter: str = ",",
reindex: bool = True
) -> pd.DataFrame:
"""Convert dataframe with columns separated by a delimiter into an
ordinary dataframe. Requires pandas 1.3.0+."""
if isinstance(columns, str):
columns = [columns]
col_dict = {
col: df[col]
.str.split(delimiter)
# Without .fillna(), .explode() will fail on empty values
.fillna({i: [np.nan] for i in df.index})
for col in columns
}
df = df.assign(**col_dict).explode(columns)
return df.reset_index(drop=True) if reindex else df
I am reading multiple JSON objects into one DataFrame. The problem is that some of the columns are lists. Also, the data is very big and because of that I cannot use the available solutions on the internet. They are very slow and memory-inefficient
Here is how my data looks like:
df = pd.DataFrame({'A': ['x1','x2','x3', 'x4'], 'B':[['v1','v2'],['v3','v4'],['v5','v6'],['v7','v8']], 'C':[['c1','c2'],['c3','c4'],['c5','c6'],['c7','c8']],'D':[['d1','d2'],['d3','d4'],['d5','d6'],['d7','d8']], 'E':[['e1','e2'],['e3','e4'],['e5','e6'],['e7','e8']]})
A B C D E
0 x1 [v1, v2] [c1, c2] [d1, d2] [e1, e2]
1 x2 [v3, v4] [c3, c4] [d3, d4] [e3, e4]
2 x3 [v5, v6] [c5, c6] [d5, d6] [e5, e6]
3 x4 [v7, v8] [c7, c8] [d7, d8] [e7, e8]
And this is the shape of my data: (441079, 12)
My desired output is:
A B C D E
0 x1 v1 c1 d1 e1
0 x1 v2 c2 d2 e2
1 x2 v3 c3 d3 e3
1 x2 v4 c4 d4 e4
.....
EDIT: After being marked as duplicate, I would like to stress on the fact that in this question I was looking for an efficient method of exploding multiple columns. Therefore the approved answer is able to explode an arbitrary number of columns on very large datasets efficiently. Something that the answers to the other question failed to do (and that was the reason I asked this question after testing those solutions).
Use set_index on A and on remaining columns apply and stack the values. All of this condensed into a single liner.
In [1253]: (df.set_index('A')
.apply(lambda x: x.apply(pd.Series).stack())
.reset_index()
.drop('level_1', 1))
Out[1253]:
A B C D E
0 x1 v1 c1 d1 e1
1 x1 v2 c2 d2 e2
2 x2 v3 c3 d3 e3
3 x2 v4 c4 d4 e4
4 x3 v5 c5 d5 e5
5 x3 v6 c6 d6 e6
6 x4 v7 c7 d7 e7
7 x4 v8 c8 d8 e8
def explode(df, lst_cols, fill_value=''):
# make sure `lst_cols` is a list
if lst_cols and not isinstance(lst_cols, list):
lst_cols = [lst_cols]
# all columns except `lst_cols`
idx_cols = df.columns.difference(lst_cols)
# calculate lengths of lists
lens = df[lst_cols[0]].str.len()
if (lens > 0).all():
# ALL lists in cells aren't empty
return pd.DataFrame({
col:np.repeat(df[col].values, df[lst_cols[0]].str.len())
for col in idx_cols
}).assign(**{col:np.concatenate(df[col].values) for col in lst_cols})
.loc[:, df.columns]
else:
# at least one list in cells is empty
return pd.DataFrame({
col:np.repeat(df[col].values, df[lst_cols[0]].str.len())
for col in idx_cols
}).assign(**{col:np.concatenate(df[col].values) for col in lst_cols})
.append(df.loc[lens==0, idx_cols]).fillna(fill_value)
.loc[:, df.columns]
Usage:
In [82]: explode(df, lst_cols=list('BCDE'))
Out[82]:
A B C D E
0 x1 v1 c1 d1 e1
1 x1 v2 c2 d2 e2
2 x2 v3 c3 d3 e3
3 x2 v4 c4 d4 e4
4 x3 v5 c5 d5 e5
5 x3 v6 c6 d6 e6
6 x4 v7 c7 d7 e7
7 x4 v8 c8 d8 e8
pandas >= 0.25
Assuming all columns have the same number of lists, you can call Series.explode on each column.
df.set_index(['A']).apply(pd.Series.explode).reset_index()
A B C D E
0 x1 v1 c1 d1 e1
1 x1 v2 c2 d2 e2
2 x2 v3 c3 d3 e3
3 x2 v4 c4 d4 e4
4 x3 v5 c5 d5 e5
5 x3 v6 c6 d6 e6
6 x4 v7 c7 d7 e7
7 x4 v8 c8 d8 e8
The idea is to set as the index all columns that must NOT be exploded first, then reset the index after.
It’s also faster.
%timeit df.set_index(['A']).apply(pd.Series.explode).reset_index()
%%timeit
(df.set_index('A')
.apply(lambda x: x.apply(pd.Series).stack())
.reset_index()
.drop('level_1', 1))
2.22 ms ± 98.6 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
9.14 ms ± 329 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
Building on @cs95’s answer, we can use an if clause in the lambda function, instead of setting all the other columns as the index. This has the following advantages:
- Preserves column order
- Lets you easily specify columns using the set you want to modify,
x.name in [...], or not modifyx.name not in [...].
df.apply(lambda x: x.explode() if x.name in ['B', 'C', 'D', 'E'] else x)
A B C D E
0 x1 v1 c1 d1 e1
0 x1 v2 c2 d2 e2
1 x2 v3 c3 d3 e3
1 x2 v4 c4 d4 e4
2 x3 v5 c5 d5 e5
2 x3 v6 c6 d6 e6
3 x4 v7 c7 d7 e7
3 x4 v8 c8 d8 e8
Here is my solution using ‘apply’ function. Main features/differences:
- offer options to specify selected multiple columns or all columns
- offer options to specify values to fill in the ‘missing’ position (through parameter fill_mode = ‘external’; ‘internal’; or ‘trim’, explanation would be long, see examples below and try yourself to change the option and check the result)
Notes: option ‘trim’ was developed for my need, out of scope for this question
def lenx(x):
return len(x) if isinstance(x,(list, tuple, np.ndarray, pd.Series)) else 1
def cell_size_equalize2(row, cols='', fill_mode='internal', fill_value=''):
jcols = [j for j,v in enumerate(row.index) if v in cols]
if len(jcols)<1:
jcols = range(len(row.index))
Ls = [lenx(x) for x in row.values]
if not Ls[:-1]==Ls[1:]:
vals = [v if isinstance(v,list) else [v] for v in row.values]
if fill_mode=='external':
vals = [[e] + [fill_value]*(max(Ls)-1) if (not j in jcols) and (isinstance(row.values[j],list))
else e + [fill_value]*(max(Ls)-lenx(e))
for j,e in enumerate(vals)]
elif fill_mode == 'internal':
vals = [[e]+[e]*(max(Ls)-1) if (not j in jcols) and (isinstance(row.values[j],list))
else e+[e[-1]]*(max(Ls)-lenx(e))
for j,e in enumerate(vals)]
else:
vals = [e[0:min(Ls)] for e in vals]
row = pd.Series(vals,index=row.index.tolist())
return row
Examples:
df=pd.DataFrame({
'a':[[1],2,3],
'b':[[4,5,7],[5,4],4],
'c':[[4,5],5,[6]]
})
print(df)
df1 = df.apply(cell_size_equalize2, cols='', fill_mode='external', fill_value = "OK", axis=1).apply(pd.Series.explode)
print('nfill_mode='external', all columns, fill_value = 'OK'n', df1)
df2 = df.apply(cell_size_equalize2, cols=['a', 'b'], fill_mode='external', fill_value = "OK", axis=1).apply(pd.Series.explode)
print('nfill_mode='external', cols = ['a', 'b'], fill_value = 'OK'n', df2)
df3 = df.apply(cell_size_equalize2, cols=['a', 'b'], fill_mode='internal', axis=1).apply(pd.Series.explode)
print('nfill_mode='internal', cols = ['a', 'b']n', df3)
df4 = df.apply(cell_size_equalize2, cols='', fill_mode='trim', axis=1).apply(pd.Series.explode)
print('nfill_mode='trim', all columnsn', df4)
Output:
a b c
0 [1] [4, 5, 7] [4, 5]
1 2 [5, 4] 5
2 3 4 [6]
fill_mode='external', all columns, fill_value = 'OK'
a b c
0 1 4 4
0 OK 5 5
0 OK 7 OK
1 2 5 5
1 OK 4 OK
2 3 4 6
fill_mode='external', cols = ['a', 'b'], fill_value = 'OK'
a b c
0 1 4 [4, 5]
0 OK 5 OK
0 OK 7 OK
1 2 5 5
1 OK 4 OK
2 3 4 6
fill_mode='internal', cols = ['a', 'b']
a b c
0 1 4 [4, 5]
0 1 5 [4, 5]
0 1 7 [4, 5]
1 2 5 5
1 2 4 5
2 3 4 6
fill_mode='trim', all columns
a b c
0 1 4 4
1 2 5 5
2 3 4 6
As of pandas 1.3.0 (What’s new in 1.3.0 (July 2, 2021)):
DataFrame.explode()now supports exploding multiple columns. Its column argument now also accepts a list of str or tuples for exploding on multiple columns at the same time (GH39240)
So now this operation is as simple as:
df.explode(['B', 'C', 'D', 'E'])
A B C D E
0 x1 v1 c1 d1 e1
0 x1 v2 c2 d2 e2
1 x2 v3 c3 d3 e3
1 x2 v4 c4 d4 e4
2 x3 v5 c5 d5 e5
2 x3 v6 c6 d6 e6
3 x4 v7 c7 d7 e7
3 x4 v8 c8 d8 e8
Or if wanting unique indexing:
df.explode(['B', 'C', 'D', 'E'], ignore_index=True)
A B C D E
0 x1 v1 c1 d1 e1
1 x1 v2 c2 d2 e2
2 x2 v3 c3 d3 e3
3 x2 v4 c4 d4 e4
4 x3 v5 c5 d5 e5
5 x3 v6 c6 d6 e6
6 x4 v7 c7 d7 e7
7 x4 v8 c8 d8 e8
Gathering all of the responses on this and other threads, here is how I do it for comma-delineated rows:
from collections.abc import Sequence
import pandas as pd
import numpy as np
def explode_by_delimiter(
df: pd.DataFrame,
columns: str | Sequence[str],
delimiter: str = ",",
reindex: bool = True
) -> pd.DataFrame:
"""Convert dataframe with columns separated by a delimiter into an
ordinary dataframe. Requires pandas 1.3.0+."""
if isinstance(columns, str):
columns = [columns]
col_dict = {
col: df[col]
.str.split(delimiter)
# Without .fillna(), .explode() will fail on empty values
.fillna({i: [np.nan] for i in df.index})
for col in columns
}
df = df.assign(**col_dict).explode(columns)
return df.reset_index(drop=True) if reindex else df