# mse: Prediction error rates In mlesnoff/rnirs: Dimension reduction, Regression and Discrimination for Chemometrics

 mse R Documentation

## Prediction error rates

### Description

Calculation and plotting of prediction error rates (RMSEP, SEP, etc. or classification error rate) for models with quantitative or qualitative responses.

Note: In function `mse`, The `R2` rate is calculated by `R2 = 1 - MSEP(fitted model) / MSEP(no model)`, where `MSEP = Sum((y_i - fit_i)^2)/n` and "no model" is the overall mean of `y`. For predictions over CV or Test sets, and/or for non linear models, it can be different from the square of the correlation coefficient between the observed values and prediction.

See examples in `plsr`, `plsda`, etc.

### Usage

``````
mse(fm, formula = ~ 1, nam = NULL, digits = 3)

err(fm, formula = ~ 1, nam = NULL, digits = 4)

plotmse(obj, namx = "ncomp", nam = "msep", group = NULL,  col = NULL,
legend = TRUE, legend.title = NULL, ncol = 1, ...)

``````

### Arguments

 `fm` An object containing at least data frames `y` (data) and `fit` (predictions). See for instance objects `y` and `fit` returned by function `plsr`. `formula` A right-hand-side formula defing the aggregation levels on which are calculated the error rates. `namx` For `plotmse`: Name (character string) of the column in `obj` to plot as x-axis. Default to `nam = "ncomp"`. `nam` For `mse`: Name (character string) of the column to consider in `fm\$y` for calulating the prediction error rate. If `NULL` (default), the last column of `fm\$y` is considered. For `plotmse`: Name (character string) of the column in `obj` to plot as y-axis. Default to `nam = "msep"`. `digits` The number of digits for the numerical outputs.

Arguments specific to `plotmse`:

 `obj` Data frame returned by `mse` or `err`. `group` A vector of length equal to the number of rows of `obj` defining the groups of rows to be plotted with different colors (default is `NULL`). `col` A color, or a vector of colors (of length equal to the number of groups), defining the color(s) of the lines representing the error rates. `legend` Only if there are groups. Logical indicationg is a legend is drawn for groups (Default to `FALSE`). `legend.title` Character string indicationg a title for the legend. `ncol` Number of columns drawn in the legend box. `...` Other arguments to pass in functions `plot` or `lines`

.

### Value

A dataframe.

mlesnoff/rnirs documentation built on April 24, 2023, 4:17 a.m.