Regression Analysis for the Anomalies!

3 次查看（过去 30 天）

Sophia 2016-11-23

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/313699-regression-analysis-for-the-anomalies

I have the ice drift anomalies(pixel value for one particular year- Avg value) data for the 36 winters; size 361*361*36

Land areas have zero ice drift
In earlier years some places has ice, so has the ice drift values but are ice free in the recent years

In my trend analysis it shows the negative trend for those areas, however i don't want to take those pixels in my analysis where there is no data for all the years. How to omit that.

%%Regression analysis for ice drift anomalies
   for i=1:361,
        for j=1:361,
           if (sum(isnan(anom_wint_r(i,j,:))> 0));
               idanom_trend_per_winter(i,j) = NaN;
            else
                stats = regstats(squeeze(anom_wint_r(i,j,:)),years','linear',{'beta' 'rsquare' 'fstat'});
                idanom_trend_per_winter(i,j) = stats.beta(2);
            end
        end
    end