[GEMPAK #MST-494826]: Linear Regression

Nick,

These are not capabilities in GEMPAK. The package I used to do this type
of research in the past was matlab or octave. Both are good at matrix
types of problems.

Steve CHiswell
Unidata User Support

> 
> Hi, I am a relatively new GEMPAK user and I am writing to see if you would be 
> able to offer any assistance - I am currently conducting an investigation on 
> long-range model bias correction on temperatures using linear regression 
> techniques.
> 
> I will be doing this by using grid analysis functions on forecast and 
> observation data, and I will be needing to compute the correlation 
> coefficient r² and also the slope and y-intercept of the regression line, y = 
> mx +b, with y being the values of the observation grid squares, and x the 
> corresponding values of the forecast grid squares.
> 
> As far as I can tell after reading the software documentation, there are no 
> functions that automatically compute these values r², m, and b, instead I 
> will need to try to compute them myself in the grid functions by using 
> formulas for these values.
> 
> I was just wondering if this is indeed true, or, are there functions 
> available that I am not aware of that will alllow me to do this more easily?
> 
> Thank you very much for your time.
> 
> 
> 
> Nick Carlisle : Development Meteorologist - 
> theweathernetwork.com<http://theweathernetwork.com/>
> contact | address@hidden<mailto:address@hidden> - 905.829.1159 x1436
> 
> 
> 
> 
> 
> 


Ticket Details
===================
Ticket ID: MST-494826
Department: Support GEMPAK
Priority: Normal
Status: Closed