How to perform operation over each dictionary in column?-CodePudding

If I have a column in dataframe with dictionaries:

col1
{10:24, 7:3}
{5:24, 1:2, 7:8}
{1:1}

How to perform operation of extracting keys from each dictionary for each rows? So I need to get:

col1
10, 7
5, 1, 7
1

How to do that? this df["col1"] = df["col1"].keys() doesnt work and I don't know why

CodePudding user response：

DataFrame has .keys() to get own indexes, not to get keys from dictionares in cells.

But you can use .apply() to run function on every elemement in column separatelly.

df['col1'] = df['col1'].apply(lambda item: item.keys())

Minimal working example:

import pandas as pd

df = pd.DataFrame({'col1':[
   {10:24, 7:3},
   {5:24, 1:2, 7:8},
   {1:1},
]})


df['col1'] = df['col1'].apply(lambda item: item.keys())

print(df)

Result (now it has tuples with numbers):

        col1
0    (10, 7)
1  (5, 1, 7)
2        (1)

BTW:

DataFrame has special method to work with strings .str which may work also with list/tuples and some of them even with dictionary

It can't get df['col1'].str.keys() because string doesn't have keys but if you use df['col1'].str[10] then you get from all dictionares elements which have key 10

0    24.0
1     NaN
2     NaN

CodePudding user response：

df["col1"] is not a dictionary - it is a tuple. That explains why you get an AttributeError. You need to iterate over each row in the dataframe column and call keys() on it.

df['col1'] = [row.keys() for row in df["col1"]]

CodePudding user response：

DataFrame.apply according to the documentation:

Apply a function along an axis of the DataFrame.

Luckily, the default axis is columns, not rows. You are going for a single column, so make your applied function check the current column.

df.apply(lambda c: c.keys() if c.name == "col1" else c)