Hosmer-Lemeshow goodness of fit test — metrics

Internal, not usually called directly

metrics_hoslem(y, yhat, g = 10, digits = c(2, 3))

Source

https://github.com/psolymos/ResourceSelection/blob/master/R/hoslem.test.R

y: Observed y, usually of the form fit$y.
yhat: Predicted y_hat, usually for the form fit$fitted
g: Number of bins to calculate quantiles.
digits: Number of decimal places of form c(2,3), where digits[1] is for chi-sq estimate and digits[2] is for p-value.

Character string of chi-sq result, df, and p-value. Significant p-value suggests poor fit.

Adapted from Peter Solymos.

fit = glm(mort_5yr~age.factor+extent.factor, data=colon_s, family="binomial")
metrics_hoslem(fit$y, fit$fitted)
#> [1] "Chi-sq(8) 1.18 (p=0.997)"