Pandas: How to assign one column value to a variable, based on another column?

Question

I have a DataFrame, for which I want to store values from the keyvalue column into a variable, based on the value of keyname.

Example DataFrame:

keyname     keyvalue
A           100,200
B           300
C           400

Expected output:

v_A = [100, 200]
v_B = 300
v_C = 400

Please add more details, so we can help you. With the current description your question is not clear. — Dani Mesejo, Nov 27 '20 at 08:33
use keyname as variable and keyvalue as value and store value in a variable — Shivam Patel, Nov 27 '20 at 08:36

S3DEV · Answer 1 · 2020-11-27T09:16:55.580

3

While this is a more verbose approach, it's posted to demonstrate the basic concept for assigning keyvalue values to a list variable, based on the value of keyname.

v_A = df.loc[df['keyname'] == 'A', 'keyvalue'].to_list()
v_B = df.loc[df['keyname'] == 'B', 'keyvalue'].to_list()
v_C = df.loc[df['keyname'] == 'C', 'keyvalue'].to_list()

Output:

['100,200']
['300']
['400']

edited Nov 27 '20 at 09:16

answered Nov 27 '20 at 08:37

S3DEV

8,768
3
31
42

SyntaxError: can't assign to literal – Shivam Patel Nov 27 '20 at 08:43
(SyntaxError in comment above does not apply to the current (edited) answer.) – S3DEV Nov 27 '20 at 08:58

jezrael · Accepted Answer · 2020-11-27T09:01:16.797

1

Close, what you need is dictionary with keys and values, because concept of strings variables in python is not recommended:

d = df.set_index('keyname')['keyvalue'].to_dict()

edited Nov 27 '20 at 09:01

answered Nov 27 '20 at 08:41

jezrael

822,522
95
1,334
1,252

TypeError: argument of type 'int' is not iterable – Shivam Patel Nov 27 '20 at 08:47
@ShivamPatel - so you need `d = df.set_index('keyname')['keyvalue'].to_dict()` only? – jezrael Nov 27 '20 at 08:50

Johnson Francis · Answer 3 · 2020-11-27T09:27:13.583

1

I can suggest two options..

Convert to a dictionary.. that will give you them as key value pairs, if that is what you want.
```
df.set_index('keyname').to_dict()
```

output:

   {'keyvalue': {'A': '100,200', 'B': '300', 'C': '400'}}

Take a transpose and you will get them in columns of dataframe and then you can convert as list

dft=df.set_index('keyname').T
v_A=list(map(int, dft['A'][0].split(",")))
v_B=list(map(int, dft['B'][0].split(",")))
v_C=list(map(int, dft['C'][0].split(",")))
print(v_A)
print(v_B)
print(v_C)

output :

   [100, 200]
   [300]
   [400]

edited Nov 27 '20 at 09:27

answered Nov 27 '20 at 09:09

Johnson Francis

249
3
17

1

I dont see why is better as culomns.. nd what that dictionary is better then jezreal answer? – adir abargil Nov 27 '20 at 09:17

Pandas: How to assign one column value to a variable, based on another column?

3 Answers3