Goodness of Fit (Modified)

Computes goodness of fit for regression model given vector or matrix of target and output values.  
   
USAGE:  
   [gf] = gfit2(t,y)  
   [gf] = gfit2(t,y,gFitMeasure)  
         
INPUT  
t: matrix or vector of target values for regression model  
   
y: matrix or vector of output from regression model.  
   
gFitMeasure: string or cell array of strings representing different forms of goodness of fit measure (eg. '1' - mean squarred error, '5' - mean absolute error, see below for details) . If gFitMeasure is not supplied, all statistics are returned as a vector. The following stats are returned:  
 
'1' - mean squared error (mse)  
'2' - normalised mean squared error (nmse)  
'3' - root mean squared error (rmse)  
'4' - normalised root mean squared error (nrmse)  
'5' - mean absolute error (mae)  
'6' - mean absolute relative error (mare)  
'7' - coefficient of correlation (r)  
'8' - coefficient of determination (d)  
'9' - coefficient of efficiency (e)  
'10' - maximum absolute error  
'11' - maximum absolute relative error  
             
OUTPUT:  
 gf: goodness of fit values between model output and target, either scalar, or vector one for each gFitMeasure chosen.

The author wishes to acknowledge the following in the creation of this submission:
Goodness of Fit