How to calculate regression coeeficients fro spatial data

Hello,

I have a £D matrix called SLP that represents sea lvel pressure values over the fram strait having the following dimensions 241x81x240 where 240 is time; and a vector representing potential temperature (TEMP) with dimension 1x240.

I wuold like to calculate the regression coefficients between the matrix and the vector to create a figure like this.

Thank you very much for your help :)

[William Rose]

@Carlotta Dentico, It will help others help you if you provide files with example data.

I do not understand the statement "I would like to calculate the regression coefficients between the matrix and the vector to create a figure like this." I do not understand the statement because

• The figure does not appear to show a regression. The figure is a colored contour plot of pressure versus geographic position.
• You have SLP(Lat,Lon,time) and TEMP(time). What regression you want to compute? Is the temperature for one specific location, or is it the mean temperature across the region, or something else? The answer to that question will help determine what regression to do.

[Mathieu NOE]

duplicate post

Correlation and regression between matrixes with NaN values - MATLAB Answers - MATLAB Central (mathworks.com)

[Carlotta Dentico]

@William Rose

• I wuold like to calculate the regression coefficients which are shown by the isoline in the attached figure.
• I want to compute the regression coefficients between the SLP in the Fram Strait (bigger area) and the mean temperature along a transect.

Hope this information can help..

[Carlotta Dentico]

@Mathieu NOE

Yes, as I wrote I found the solution for the correlation, but I'm still working on the regression coefficient.

[William Rose]

I see that you have received excellent help from @dpb and @MarKf in response to your closely-related posting. As a courtesy to them, I recommend that you accept one of their answers.

You said above:

• I wuold like to calculate the regression coefficients which are shown by the isoline in the attached figure.
• I want to compute the regression coefficients between the SLP in the Fram Strait (bigger area) and the mean temperature along a transect.

What does "the regression coefficients which are shown by the isoline" mean? An isoline does not show regression coefficients. I assume the isolines above are isobars, at one moment in time, because you said you have pressure at 81x241 locations, and at 240 times. Please clarify.

When you say "I want to compute the regression coefficients between the SLP in the Fram Strait (bigger area) and the mean temperature along a transect.", is this a different set of regression coefficients than the ones "shown by the isoline"? Or is this statement another reference to the same regression coefficients? Do you have SLP along a transect, and temp along the same transect? If so, then you have two vectors of equal length, and the regression coefficient between them is easily computed.

You said "mean temperature along a transect". Your original post said the Temp vector is 1x240, and you have 240 time points. Therefore I assume you have a single temperature at each time, not temperature at each point along a transect. Do I assume correctly?

If you want to find a multiple regresion between pressures across space and time, and temperature at each time, then explain how you want to parition spatial pressure data at each time point.

As I requested previously: provide data please. You provided two data files at your other post. Each has dimensions 241x81. Neither of them have the dimensions you referred to in this posting: 241x81x240 and 1x240.