replace nan with a value

5 views (last 30 days)
Daphne Mariaravi
Daphne Mariaravi on 21 Jun 2017
Commented: Star Strider on 21 Jun 2017
I have a column in which some values are missing and shown as NaN. I need to replace all these NaN with the median of the column. Here is what I tried: I am extracting column 4 in which there are missing values, find the median value, Use isnan to replace the logical 1 with the median.
Column 4 values: 1 2 3 4 5 6 NaN 12 10 NaN 4 5 NaN
C=num(:,4);
median=nanmedian(C);
R(isnan(C))=median;
But with this I am getting results as : 0 0 0 0 0 0 12 0 0 12 0 0 12
But the expected result is: 1 2 3 4 5 6 12 12 10 12 4 5 12
Any suggestions ?

Accepted Answer

Star Strider
Star Strider on 21 Jun 2017
Please do not use ‘median’ as a variable name.
I don’t know where the ‘12’ value comes from, since the median of the vector you posted is 4.5.
Other than that, ‘R’ does not magickally become ‘C’. You have to create it first.
This works for me:
C = [1 2 3 4 5 6 NaN 12 10 NaN 4 5 NaN]';
median_4 = nanmedian(C);
R = C;
R(isnan(C)) = median_4;
  2 Comments
Daphne Mariaravi
Daphne Mariaravi on 21 Jun 2017
The median value 12 wasn't actually calculated. I just filled in some random values just to give a picture of what my problem was. Sorry about that. The code works! Thanks
Star Strider
Star Strider on 21 Jun 2017
My pleasure!

Sign in to comment.

More Answers (1)

Steven Lord
Steven Lord on 21 Jun 2017
If you're using release R2016b or later use fillmissing.
z = [1 2 3 4 5 6 NaN 12 10 NaN 4 5 NaN].';
y = fillmissing(z, 'constant', median(z, 'omitnan'));
showBothColumnsSideBySide = [z, y]
  3 Comments
Jan
Jan on 21 Jun 2017
@Daphne: This happens, because "median" was defined as a variable on your computer. Do not shadow builtin functions by variables. Solution: Restart Matlab or:
clear median
Daphne Mariaravi
Daphne Mariaravi on 21 Jun 2017
Yes. Got it! Thank you

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!