I have a small data set. I tested Lowess using two different bandwidths, 0.6 and 0.8.
Is there a way to judge which bandwidth is better explaining the relationship between V1 and V2? I understand that this is not a model fitting procedure.
Also, can I pull the mathematical equations so I can see how these graphs are drawn? any return values?
regards,
Chul
Is there a way to judge which bandwidth is better explaining the relationship between V1 and V2? I understand that this is not a model fitting procedure.
Also, can I pull the mathematical equations so I can see how these graphs are drawn? any return values?
regards,
Chul
Code:
* Example generated by -dataex-. For more info, type help dataex clear input float(V1 V2) .7829959 .4631579 .7560192 .4502408 .7296887 .4381169 .7516532 .4533431 .7509184 .4761205 .7464678 .4764521 .7682927 .4784792 .778659 .4996395 .7199414 .3533724 end
HTML Code:
lowess V2 V1, bwidth(0.6) // bandwidth = 0.6 lowess V2 V1, bwidth(0.8) // bandwidth = 0.8
Comment