Random value for each row in pandas data Frame

Question

Hi I have the following data frames:

import numpy as np
import pandas as pd
df = pd.DataFrame()
df['T1'] = ['A','B','C','D','E']
df['T2'] = ['G','H','I','J','K']
df['Match'] = df['T1'] +' Vs '+ df['T2']
Nsims = 5
df1 = pd.DataFrame((pd.np.tile(df,(Nsims,1))))

I created two new columns T1_point and T2_point by summing of five random numbers. when I do as follow: it gave me the same number for all rows.

Ninit = 5
df1['T1_point'] = np.sum(np.random.uniform(size=Ninit))
df1['T2_point'] = np.sum(np.random.uniform(size=Ninit))

What I wanted to do is that I would like to get different values for each row by using random number.

How could I do that?

Thanks

Zep.

Why are you summing them if you want different values? This makes no sense? — cs95
– cs95, Commented Jun 27, 2018 at 4:30
@coldspeed, I am trying to replicate the penalty shoot process.That's why I used 5 random number and sum it so that i can decide who will win — Zephyr
– Zephyr, Commented Jun 27, 2018 at 4:35
I did not understand. So you want to take 25 numbers and sum them in groups of 5? — cs95
– cs95, Commented Jun 27, 2018 at 4:38

ajsp · Accepted Answer · 2018-07-12 17:22:10Z

11

What you are basically asking is for a random number in each row. Just create a list of random numbers then and append them to your dataframe?

import random

df1['RAND'] = [ random.randint(1,10000000)  for k in df1.index]

print df1

    0  1     RAND
0   A  G  6850189
1   B  H  3692984
2   C  I  8062507
3   D  J  6156287
4   E  K  7037728
5   A  G  7641046
6   B  H  1884503
7   C  I  7887030
8   D  J  4089507
9   E  K  4253742
10  A  G  8947290
11  B  H  8634259
12  C  I  7172269
13  D  J  4906697
14  E  K  7040624
15  A  G  4702362
16  B  H  5267067
17  C  I  3282320
18  D  J  6185152
19  E  K  9335186
20  A  G  3448703
21  B  H  6039862
22  C  I  9884632
23  D  J  4846228
24  E  K  5510052

edited Jul 12, 2018 at 17:22

answered Jun 27, 2018 at 4:45

ajsp

2,69025 silver badges35 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

letmecheck Over a year ago

random.randint() doesn't guarantee unique values always

ajsp Over a year ago

@Mohan Babu, then he could just use a range, or even use the index for uniqueness. I dont really understand what he is doing to be honest.

Zephyr Over a year ago

Thanks ajsp. I could extend the process using your advice.

Mark Andersen · Accepted Answer · 2023-02-23 23:50:52Z

2

The list comprehension answer is not optimal.

A simpler and more efficient answer is to use the size parameter available in numpy which creates a matching array:

import numpy as np
df1['RAND'] = np.random.randint(1,10000000, size=len(df1))

answered Feb 23, 2023 at 23:50

Mark Andersen

1,0489 silver badges8 bronze badges

Collectives™ on Stack Overflow

Random value for each row in pandas data Frame

2 Answers 2

3 Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related