Converting dictionary of tuples and lists as key values to pandas dataframe

user_17175416 0
Viewed 102

I have a dictionary with tuples as keys and values as lists. I would like to convert it to a pandas dataframe for exporting it as an Excel table later.

Could someone suggest a clean way to achieve the table as below? I would like to have the elements in the list in separate columns.

dict={(a,b):[1,2][4,5],(a,c):[7,8][1,3],(b,c):[1,8][1,3]}

K X Y
a,b 1 2
a,b 4 5
a,c 7 8
a,c 1 3
b,c 1 8
b,c 1 3

Tried this, not getting desired result: df=pd.DataFrame.from_dict(dict, orient='index').T

python-3.6 pandas dictionary python
2 Answers

Use:

import pandas as pd

# toy data
dic = {("a", "b"): [[1, 2], [4, 5]], ("a", "c"): [[7, 8], [1, 3]], ("b", "c"): [[1, 8], [1, 3]]}

# un-ravel the data to create the rows of the DataFrame
data = [[key, *value] for key, values in dic.items() for value in values]

# actually create the DataFrame
df = pd.DataFrame(data, columns=["K", "X", "Y"])
print(df)

Output

        K  X  Y
0  (a, b)  1  2
1  (a, b)  4  5
2  (a, c)  7  8
3  (a, c)  1  3
4  (b, c)  1  8
5  (b, c)  1  3

The expression:

data = [[key, *value] for key, values in dic.items() for value in values]

is a list comprehension, is equivalent to:

data = []
for key, values in dic.items():
    for value in values:
        data.append([key, *value])
user_40015920

Check if this helps:

import pandas as pd
dict={('a','b'):[[1,2],[4,5]],('a','c'):[[7,8],[1,3]],('b','c'):[[1,8],[1,3]]}
dt = []
for k,v in dict.items():
    for ls in v:
        dt.append([k,ls[0],ls[1]])

df = pd.DataFrame(dt,columns = ['Key','x','y'])

print (df)
user_171748860
Related