My data sheet has 250 columns with column names '1'...'250'. After random sampling each column, I want to save all the sampled columns in a single csv file with same column names as in the data sheet.
for i in range(1,250):
z=np.random.choice(df[i], len(df), replace=False)
How to proceed?
I believe need assign back each column and then write to csv by to_csv:
for i in range(1,250):
df[i]=np.random.choice(df[i], len(df), replace=False)
df.to_csv(file, index=False)
Another solution with sample:
df1 = df.sample(len(df.columns), replace=False)
df.to_csv(file, index=False)
Sample:
df = pd.DataFrame(np.random.randint(30, size=(5,4))).rename(columns=lambda x: x+1)
print (df)
1 2 3 4
0 12 10 2 14
1 9 14 28 4
2 9 11 14 8
3 22 8 2 9
4 28 3 23 6
df1 = df.sample(len(df.columns), axis=0, replace=False)
print (df1)
1 2 3 4
3 22 8 2 9
0 12 10 2 14
2 9 11 14 8
4 28 3 23 6
for i in range(1,5):
df[i]=np.random.choice(df[i], len(df), replace=False)
print (df)
1 2 3 4
0 28 11 23 8
1 22 3 2 14
2 9 10 28 4
3 9 8 14 9
4 12 14 2 6
