Asked by Mohammad
on 5 Dec 2012

I have huge number of precipitation daily data in a column (say for 60 years) for 400 meteorological stations. Now I need to convert it to monthly. How can I do that? thanks.

*No products are associated with this question.*

Answer by Babak
on 5 Dec 2012

Every 30 data rows, I would average them out and save it in 1 row.This way, 30 days shrinks down to 1 data (a month). Or another way, you can simply delete 29 data rows and keep only one from every 30 data rows.

Answer by Andrei Bobrov
on 5 Dec 2012

Edited by Andrei Bobrov
on 5 Dec 2012

one way

M - your data matrix with 4 columns < Year Month Day data > - double type

[a,~,c] = unique(M(:,1:2),'rows'); out = [a, accumarray(c,M(:,4),[],@mean)];

Answer by per isakson
on 5 Dec 2012

Edited by per isakson
on 5 Dec 2012

Given thirty days in every month try this

%% daily_val = rand( 30*12, 1 ); %% daily_val = reshape( daily_val, 30, [] ); %% monthly_val = sum( daily_val, 1 );

here I assume that the first value in daily_val is the first value of a month. This is intended as a start.

Answer by Chad Greene
on 5 Nov 2014

`downsample_ts` can do this by

precip_monthlymean = downsample_ts(precip_daily,t_daily);

Opportunities for recent engineering grads.

## 2 Comments

## per isakson (view profile)

Direct link to this comment:http://www.mathworks.com/matlabcentral/answers/55819#comment_115536

Months have 28, 29, 30 or 31 days, which makes it a bit tricky.

Do you have a table with

rather than just a column?

Does the length of the months matter?

## Mohammad (view profile)

Direct link to this comment:http://www.mathworks.com/matlabcentral/answers/55819#comment_115540

my data format is Year Month Day data

that is in 4 column. Length of month is always 30. Does it help for providing your answer? thanks.