Joint density classification

Fits a generic joint density model for the predictors and class variables.

jdify(formula, data, jd_method = "cctools", ...)

Arguments

formula	an object of class "formula"; same as `stats::lm()`.
data	matrix, data frame, list or environment (or object coercible by `base::as.data.frame()`) containing the variables in the model.
jd_method	an object of class `"jd_method"` or a character string for built-in methods, see `jd_method()`.
...	additional parameters passed to `fit_fun`.

Value

An object of class jdify.

Details

For predicting the class or conditional probabilities, see, predict.jdify().

References

Nagler, T. (2017). A generic approach to nonparametric function estimation with mixed data. arXiv:1704.07457

Examples

# simulate training and test data
dat <- data.frame(
    cl = as.factor(rbinom(10, 1, 0.5)),
    x1 = rnorm(10),
    x2 = ordered(rbinom(10, 1, 0.3), 0:1)
)

model <- jdify(cl ~ x1 + x2, dat)              # fitted joint density
pred <- predict(model, dat)                    # class predictions
probs <- predict(model, dat, what = "probs")   # conditional probabilities

# in-sample performance
assess_clsfyr(probs[, 1], dat[, 1] == 0, measure = c("ACC", "F1"))
#>    threshold measure     value
#> 1        0.0     ACC 0.7000000
#> 2        0.1     ACC 0.7000000
#> 3        0.2     ACC 0.7000000
#> 4        0.3     ACC 0.7000000
#> 5        0.4     ACC 0.7000000
#> 6        0.5     ACC 0.8000000
#> 7        0.6     ACC 0.9000000
#> 8        0.7     ACC 0.4000000
#> 9        0.8     ACC 0.4000000
#> 10       0.9     ACC 0.4000000
#> 11       1.0     ACC 0.3000000
#> 12       0.0      F1 0.8235294
#> 13       0.1      F1 0.8235294
#> 14       0.2      F1 0.8235294
#> 15       0.3      F1 0.8235294
#> 16       0.4      F1 0.8235294
#> 17       0.5      F1 0.8750000
#> 18       0.6      F1 0.9333333
#> 19       0.7      F1 0.2500000
#> 20       0.8      F1 0.2500000
#> 21       0.9      F1 0.2500000
#> 22       1.0      F1 0.0000000