Writing a loop to calculate a seasonal cycle and then plot the seasonal cycle?

My current code is this:

{for mdo=1:12

index_month = find(month==mdo);

month_mean = mean(avhrr.sst(index_month));

seasonal_cycle(mdo)=month_mean;

s_cycle_month(index_month) = month_mean;

end

hold on

plot(t, month_mean, 'r')}

When I run this, it produces a blank graph. I think the loop is correctly producing the seasonal cycle, how can I get it to plot the graph?

[bushra raza]

Hi Victoria,

are you done with seasonal cycle calculations?

i have 30 years hourly data , and i want to see the seasonal cycle, and trend etc and further analysis. please guide me for its start?

Regards,

BNA

Answer 1

[Cedric Wannaz]

Where is t defined? In addition, month_mean is your intermediary variable; you want to plot either seasonal_cycle or s_cycle_month, which are your output vectors.

[Victoria Woods]

t is defined earlier on in my code. I've changed my plot to t against s_cycle_month and it doesn't even open a figure anymore.

[Cedric Wannaz]

Check every variable. Maybe one is an empty array. Try to compute intermediary terms manually, e.g.

 index_month = find(month==1)

If it outputs nothing it means that month, which you suppose is a vector with some elements equal to 1, has in fact no element equal to 1. If the output looks fine, then evaluate:

 avhrr.sst(index_month)

If the output is not a vector of floating point elements, then maybe avhrr.sst is not what you think it is.

Check every expression for a given month index at first, so you are sure that the inputs are what you think they are, and that the outputs are valid. Once this works, embed it in the loop and try to debug the loop.

Also, try plotting seasonal_cycle first with e.g.

 plot( seasonal_cycle ) ;

If you tried plotting s_cycle_month against t, I guess that it means that t is not 1:12. This is a little more complex to debug, because s_cycle_month is more complex than seasonal_cycle (you build the former by re-injecting the mean at positions that correspond to the current month apparently).

Answer 2

[Chad Greene]

Check out the season and climatology functions in the Climate Data Toolbox for Matlab. The functions make it easy to extract seasonal cycles, (and they properly detrend the data before the calculation).