fill missing values inside a dataset that contains nan

Hi all,

I have a dataset of some weather variables such as rainfall and solar radiation. But there are some missing values in the database. I am clueless on how to fill the missing values using interpolation method.

I cannot delete the entire row with Nans since it will affect the overall results.

I have tried to refer to some examples but still not working my dataset is called mersing which contains 1029 rows and 8 columns. The weather data start from column 5 up to column 8.

the dataset looks like this( station no, year, day, rainfall, windspeed, solarradiation, evaporation)

Here are sample of my code that I am currently working on which based on answer provided by Sven but still not working.

load mersing_data fulldata_mersing = mersing; for c = 5:size(mersing,2) % start loop from column 5 until all column nanRows = isnan(mersing(:,c)); nanRows = fullData_mersing(nanRows,c) == interp1(mersing(~nanRows,1), mersing(~nanRows,c), mersing(nanRows,1)); fulldata_mersing(nanRows,c) = interp1(mersing(~nanRows,1), mersing(~nanRows,c), mersing(nanRows,1), 'nearest','extrap'); end

Answer 1

[Image Analyst]

Give John D'Errico's inpaint_nans a try: http://www.mathworks.com/matlabcentral/fileexchange/4551-inpaintnans

[Reyasudin]

The John D'Errico's inpaint_nans solved my problems...tq for the info