For example
df = {'a': (1,2,3)}
pd.DataFrame(df)
and I want to copy the column for several times, and put them in to the same dataframe. I hope it finally becomes
a b c
0 1 1 1
1 2 2 2
2 3 3 3
Codes I tried need manually name the columns. And I cannot build a loop to auto name columns.
If you do not want to write by hand the alphabet, you can do:
import pandas as pd
df = {'a': (1,2,3)}
dd = pd.DataFrame(df)
numcopies = 4 #how many copies you want of the original column named 'a'
names = [chr(ord('a') + i) for i in range(1, numcopies+1)]
for n in names:
dd[n] = dd['a']
chr() and ord() are built-in functions which comes in handy to generate sequence of letters / characters.
A more concise version of it (with no need to store the names in a variable) is:
for i in range(1, numcopies+1):
dd[chr(ord('a') + i)] = dd['a']
If you want to have a lot of copied columns, have a look at this answer. It is a good way to produce sequences of strings longer that 1 character with alphabet letters only based on an input number (so you do not end with strange symbols as name column).
You can do:
for i in range(1, numcopies+1):
dd[colnum_string(i)] = dd['a']
where column_string is the function defined in the link above.
You could use string.ascii_lowercase and DataFrame.assign for an efficient way of doing this:
from string import ascii_lowercase
df = {'a': (1,2,3)}
df = pd.DataFrame(df)
# where n is the number of columns you want
n = 5
df_new = df.assign(**{x: df.iloc[:, 0] for x in ascii_lowercase[:n]})
print(df_new)
[output]
a b c d e
0 1 1 1 1 1
1 2 2 2 2 2
2 3 3 3 3 3
You can do something like this:
names = ['b', 'c', 'd', 'e'] #names for the columns
for i in names:
df[i] = df['a'] #assigns the columns in a loop
