# a3.base: Base A3 Results Calculation In A3: Accurate, Adaptable, and Accessible Error Metrics for Predictive Models

## Description

This function calculates the A3 results. Generally this function is not called directly. It is simpler to use `a3` (for arbitrary models) or `a3.lm` (specifically for linear regressions).

## Usage

 ```1 2 3``` ```a3.base(formula, data, model.fn, simulate.fn, n.folds = 10, data.generating.fn = replicate(ncol(x), a3.gen.default), p.acc = 0.01, features = TRUE, slope.sample = NULL, slope.displacement = 1) ```

## Arguments

 `formula` the regression formula. `data` a data frame containing the data to be used in the model fit. `model.fn` function used to generate a model. `simulate.fn` function used to create the model and generate predictions. `n.folds` the number of folds used for cross-validation. Set to 0 to use Leave One Out Cross Validation. `data.generating.fn` the function used to generate stochastic noise for calculation of exact p values. `p.acc` the desired accuracy for the calculation of exact p values. The entire calculation process will be repeated 1/p.acc times so this can have a dramatic affect on time required. Set to `NULL` to disable the calculation of p values. `features` whether to calculate the average slopes, added R^2 and p values for each of the features in addition to the overall model. `slope.sample` if not NULL the sample size for use to calculate the average slopes (useful for very large data sets). `slope.displacement` the amount of displacement to take in calculating the slopes. May be a single number in which case the same slope is applied to all features. May also be a named vector where there is a name for each feature.

## Value

S3 `A3` object containing:

 `model.R2` The cross validated R^2 for the entire model. `feature.R2` The cross validated R^2's for the features (if calculated). `model.p` The p value for the entire model (if calculated). `feature.p` The p value for the features (if calculated). `all.R2` The R^2's for the model features, and any stochastic simulations for calculating exact p values. `observed` The observed response for each observation. `predicted` The predicted response for each observation. `slopes` Average slopes for each of the features (if calculated). `all.slopes` Slopes for each of the observations for each of the features (if calculated). `table` The A3 results table.

A3 documentation built on May 30, 2017, 12:41 a.m.