Replacing NaN with last real value

Hi folks, I have hundreds of excel files in .csv format that I have to "clean" getting rid of all teh NaN values and replacin them by the last value.

I have tryied using the repnan function but somethinf seems wrong.

Here is what i did

xlsread ECUSIMTEST0000-000-01.csv
for k=1:size(ans,2)
    ans(:,k) = repnan(ans(:,k),"previous")
end

Where ans is the double imported from excel. I am a beginner in matlab so no step is trivial to me.

Thanks !!!

Answer 1

[Star Strider]

If you have R2016b or later releases, use the fillmissing function.

I am not certain what you want to do, so see the method and EndValues documentation sections to fill the NaN values with the values you want. The 'previous', 'next', or 'nearest' methods would llikely be appropriate.

[Diego Dranuta]

Thank you! This worked out

[Star Strider]

As always, my pleasure!

[Asliddin Komilov]

I have set of 25 variables (matrices), can I have repliced all zero value to NAN in all of the at once, without doing it for each of them separately?

thanks