topic Re: Lack of Fit is blank in SAS Programming

Lack of Fit is blank

samer_badri75 — Fri, 15 Feb 2019 02:55:39 GMT

data d;
   input Day Temp RH;
   label Day = "Reaction Day (Hours)"
         Temp = "Temperature (Degrees Centigrade)"
         RH  = "Percent of humidity";
   datalines;
15	24	78
12	24	79
15	24	78
11	24	79
13	22	75
14	22	75
14	22	75
13	22	75
9	30	74
10	30	74
10	30	74
10	30	74
11	26	62
10	25	62
10	26	62
11	26	62





;
ods graphics on;
proc rsreg data=d plots=(ridge surface);
   model Day=RH Temp / lackfit;
   ridge max;
run;

Greeting All

I used the SAS code got blank for Lack of Fit, Is there something wrong with the data? If I got a lack of fit significant what is the other method to get a better model.

Thanks in advance

Re: Lack of Fit is blank

mkeintz — Fri, 15 Feb 2019 08:00:47 GMT

I think it's blank because the degrees of freedom attributable to lack-of-fit is zero.

According to the wikipedia article Lack-of-fit sum of squares the degrees of freedom for lack of fit is n-p, where

n is the number of levels of independent variables (you have 6 such levels - i.e. 6 combinations of RH and TEMP).
p is the number of parameters. I see your model allocates 5 degrees of freedom to the model. Since rsreg shows zero df for lack-of-fit, I guess it adds 1 df for an intercept term - yielding p=6.

Re: Lack of Fit is blank

samer_badri75 — Fri, 15 Feb 2019 15:02:06 GMT

Many thanks
I will try that
Have a great day