for loop to create a new column containing the mean of specific columns and generating a "missing values" message where NaN

user_13223827 0
Viewed 144

I am very very new to python so apologies in advance, i have been trying to solve this for hours.

I have a set of data like this, but much much larger:

  A   B   C   D   
1 23  16  NaN 14
2 26  17  23  23
3 23  NaN 22  25
4 24  34  28  28

I need to make another column (E) that takes the average score of specific columns, such as B,C and D, for that particular row.

If across this row, there are any missing values (NaN) then i need "missing data" to be shown in row E in place of where the mean score would be.

I have tried changing the NaN outputs to 0s (which was successful) and then running something similar to the following (my code has become a complete mess and i have forgotten where i started or what i have tried):

composite = []
for df in column ["A","B","C"]:
    if value > 0:
        composite.append(df[:, ["A","B","C"]].mean(axis=1))
    else:
        composite.append("missing value(s)")

df["composite"] = composite
print(df)

I know that this code here probably has lots of errors but its a rough structure of what i am trying to do.

I have also tried every method i could find on google, including other techniques such as .loc functions. I did not want to have to ask for help as usually i'm confident i can find the solution on my own using previously posted questions, but in this instance i haven't been able to make anything work despite spending hours combing through google.

Any help would really be appreciated. I have also been told that i HAVE to use a for loop. Additionally, if this can be done without changing the NaN values to 0 this would be preferable.

Thank you for any input.

pandas for-loop python
1 Answers

I believe you could probably attempt something like this:

# Create column E with missing values
df["E"] = "missing values"

for idx, row in df.iterrows():
    # Try to convert your values as foat, in case any of them is NaN it will thow
    # an exception and pass
    try:
        B = float(df.iloc[idx,"B"])
        C = float(df.iloc[idx,"C"])
        D = float(df.iloc[idx,"D"])
        df.iloc[idx,"E"] = (B + C + D)/3  # Calculate the mean value and place it on E
    except:
        pass
user_63934760
Related