PROC PLS RESULTADOS PREDITOS IGUAIS

[deysealves]

O que você está fazendo? a analise é para predizer valores de um composto em amostras. Você tem os valores preditos adicionados todos iguais. Alguém pode me ajudar?

porque os valores aparecem a igual e que fazem para corrigi-los.

obrigada..

PROC PLS DATA = composto MÉTODO = pls nfac = 12 cv = bloco (7) PLOTS = (diagnósticos); 
MODELO COMPOSTO = a1-a1401; 
output out = preditos2 p = p_composto; 
corre; 
ods gráficos desativados; 

proc print data = preditos2; 
var p_composto composto; 
corre;

 

[PaigeMiller]

From Google Translate:

Hi, I'm trying to do some analysis using these commands below. the analysis is for predicting values of a compound in samples. However, the predicted values are all the same. Can someone help me?

they know why these values appear the same and what I do to correct them.

thanks .

It's possible that the cross-validation determined that the best fitting model was the mean of the Y values, and that no other fit predicted as well. (In other words, it's all noise, there's no structure that can be used to predict).

Show us the cross-validation part of the output of PROC PLS. Click on the {i} icon and then copy and paste the cross-validation into that window.

[PaigeMiller]

Adding: one reason you might get this constant prediction is if you have a serious outlier(s) in your data, this may result in the conclusion that you can't predict anything, and so removing the outlier(s) may allow predictability.

[Kurt_Bremser]

When posting code, don't run it through a translator. SAS is only one uniform language all over the world 😉