How to reference Discontiguous data rows in a function? -from new user

1 vue (au cours des 30 derniers jours)
Justin
Justin le 22 Sep 2014
Commenté : Justin le 22 Sep 2014
I have imported a data set to model a decision tree using fitctree function.
My overall aim is to model a training sample from the data, and predict the remainder as out-of-sample - but I want to determine out-of-sample prediction for the entire dataset, so my approach is to model 4/5 of data as training (say, rows 1-800), and record prediction for the other 1/5 (rows 801-1000).
My problem is that I then want the fitctree fucntion to use rows 1-600 AND rows 801-1000 as the training sample, so that I can record tree.predict for rows 601-800, and so on until I can record predicted values for 5 sets of test data to populate results for the entire 1000 row set.
My firt function looked similar to this: fitctree(T(1:800,2:end),T(1:800,1)) X values are in columns 2:end, and the Y values are in column 1.
Any tips would be great! Thanks!

Connectez-vous pour répondre à cette question.

Réponse acceptée

Matt J
Matt J le 22 Sep 2014
Something like this, perhaps,
rows=[1:600,801:1000];
fitctree(T(rows,2:end),T(rows,1))

Plus de réponses (0)

Catégories

En savoir plus sur Classification Trees dans Help Center et File Exchange

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by