mdl = fitglm(X, y)
mdl = fitglm(X, y, modelspec)
mdl = fitglm(X, y, ..., Name, Value)
yfit = predict(mdl, Xnew)
fitglm creates a GeneralizedLinearModel object from numeric predictors X and response y.
Supported distributions are normal, binomial, and poisson. Supported links are identity, log, and logit, with canonical defaults for each distribution.
Supported model specifications include constant, linear, interactions, quadratic, purequadratic, and numeric term matrices. Name-value arguments include Distribution, Link, Intercept, PredictorNames, ResponseName, MaxIter, and TolFun.
X = [0; 1; 2; 3; 4; 5; 6; 7];
y = [1; 1; 2; 3; 5; 8; 13; 21];
mdl = fitglm(X, y, 'Distribution', 'poisson');
yfit = predict(mdl, [2; 4; 6])| Version | Description |
|---|---|
| 2.0.0 | initial version |