How to cycle through values in Pandas Dataframe?

Question:

What’s the fastest way to cycle through and repeat names in one dataframe when updating certain rows in a second dataframe? Thank you for your help. I’ve spent many days on this and am getting nowhere.

Rows with Rules 4,5,6,10,11,12 should cycle through and repeat names in the Names Dataframe.

This is the dataframe I want to change:

Rule Name
12 Nan
9 Nan
7 Nan
1 Nan
3 Nan
2 Nan
4 Nan
6 Nan
9 Nan
4 Nan
1 Nan
4 Nan
5 Nan
2 Nan
11 Nan
6 Nan
9 Nan
2 Nan
10 Nan
3 Nan
10 Nan
3 Nan
1 Nan
4 Nan
4 Nan
3 Nan
7 Nan
7 Nan
12 Nan

This is the result I am trying to achieve:

Rule Name
12 Adams
9 Nan
7 Nan
1 Nan
3 Nan
2 Nan
4 Baker
6 Clark
9 Nan
4 Davis
1 Nan
4 Adams
5 Baker
2 Nan
11 Clark
6 Davis
9 Nan
2 Nan
10 Adams
3 Nan
10 Baker
3 Nan
1 Nan
4 Clark
4 Davis
3 Nan
7 Nan
7 Nan
12 Adams

Name Dataframe

Name
Adams
Baker
Clark
Davis

I tried to numpy vectorize, but only got one name for all the entries that I wanted to change.

Asked By: user1574881

||
Source

Answers:

IIUC, you can

  1. check if each element of df1 Rule column isin your rule list.
  2. Then cycle through and repeat names of df2 Name column depending on the length of the contained rule.
  3. At last conditionally assign the repeat names to df1 Rule column.
lst = [4,5,6,10,11,12]
m = df1['Rule'].isin(lst)
df1.loc[m, 'Name'] = (df2['Name'].tolist() * (m.sum()//len(df2['Name']) + 1))[:m.sum()]

print(df1)

    Rule   Name
0     12  Adams
1      9    Nan
2      7    Nan
3      1    Nan
4      3    Nan
5      2    Nan
6      4  Baker
7      6  Clark
8      9    Nan
9      4  Davis
10     1    Nan
11     4  Adams
12     5  Baker
13     2    Nan
14    11  Clark
15     6  Davis
16     9    Nan
17     2    Nan
18    10  Adams
19     3    Nan
20    10  Baker
21     3    Nan
22     1    Nan
23     4  Clark
24     4  Davis
25     3    Nan
26     7    Nan
27     7    Nan
28    12  Adams
Answered By: Ynjxsjmh

One option is to use modulo

l = [4,5,6,10,11,12]
m = df['Rule'].isin(l)

m.cumsum().sub(1).mod(len(df2['Name'])).map(dict(enumerate(df2['Name']))).where(m)

Output:

0     Adams
1       NaN
2       NaN
3       NaN
4       NaN
5       NaN
6     Baker
7     Clark
8       NaN
9     Davis
10      NaN
11    Adams
12    Baker
13      NaN
14    Clark
15    Davis
16      NaN
17      NaN
18    Adams
19      NaN
20    Baker
21      NaN
22      NaN
23    Clark
24    Davis
...
25      NaN
26      NaN
27      NaN
28    Adams
Answered By: rhug123