Raw Message

Message-ID: <1447611546.67830.YahooMailAndroidMobile@web181302.mail.ne1.yahoo.com>
Date: 2015-11-15T18:19:06Z
From: Seyed Jalil Alavi
Subject: Relative mean error in kriging

Dear All

I am comparing kriging and IDW methods in mapping Forest Site Productivity using 10-fold cross validation.

Both kriging and IDW methods produced negative mean error. Now I want to use relative mean error for comparing these methods other than RMSE and mean absolute error. here is the results of kriging;

mean of response variable: 34.76982

mean error: -0.03613827

mean absolute error: 1.598008

RMSE: 2.053376

how can I calculate relative mean error?

I read somewhere we can use this function for calculating relative mean error:

OK_CV <- krige.cv(Site_Form ~1, ~X+Y, Data, model = model1.out, nfold=10)

# mean error, ideally 0:

ME_OK <- mean(OK_CV$observed - OK_CV$ var1.pred) ME_OK

### Mean Absolutely Error MAE_OK <-mean(abs(OK_CV$residual)) MAE_OK

### Relative Mean Error MEr_OK <- (ME_OK/mean(Data$Site_Form))*100 MEr_OK

### RMSE RMSE_OK <-sqrt(mean(OK_CV$residual^2)) RMSE_OK

### Relative RMSE

RMSEr_OK <- (RMSE_OK/mean(Data$Site_Form))*100 RMSEr_OK

if I use the above function for relative mean error, the result will be negative!!!

How can I interpret the negative value?

I would be very grateful if anyone can help me to calculate relative mean error in R.

Regards
Jalil
Sent from Yahoo Mail on Android


	[[alternative HTML version deleted]]