Movatterモバイル変換

Type:

Package

Title:

Tidy Prediction and Plotting of Generalised Additive Models

Version:

1.0.0

Date:

2024-12-18

Description:

Provides functions that compute predictions from Generalised Additive Models (GAMs) fitted with 'mgcv' and return them as a tibble. These can be plotted with a generic plot()-method that uses 'ggplot2' or plotted as any other data frame. The main function is predict_gam().

License:

MIT + file LICENSE

URL:

https://github.com/stefanocoretta/tidygam,https://stefanocoretta.github.io/tidygam/

BugReports:

https://github.com/stefanocoretta/tidygam/issues

Encoding:

UTF-8

LazyData:

true

Imports:

cli, dplyr, ggplot2, glue, insight, magrittr, mgcv, rlang,stringr, tibble, tidyr, tidyselect

Suggests:

knitr, rmarkdown

VignetteBuilder:

knitr

Language:

en-US

RoxygenNote:

7.3.2

Depends:

R (≥ 2.10)

NeedsCompilation:

Packaged:

2024-12-18 15:32:50 UTC; ste

Author:

Stefano Coretta [aut, cre]

Maintainer:

Stefano Coretta <stefano.coretta@gmail.com>

Repository:

CRAN

Date/Publication:

2024-12-18 15:50:02 UTC

Pipe operator

Description

Seemagrittr::%>% for details.

Usage

lhs %>% rhs

Arguments

lhs

A value or the magrittr placeholder.

rhs

A function call using the magrittr semantics.

Value

The result of callingrhs(lhs).

Number of gestures by infants at 10, 11 and 12 months

Description

This data table contains counts of three type of gestures performed by 60infants from Bengali, Chinese and British backgrounds.

Usage

gest

Format

A tibble with 540 observations and 5 variables:

dyad: Unique parent/infant dyad ID.
background: Cultural background of dyad.
months: Time point in infant months.
gesture: Type of gesture.
count: Number of gestures.

Source

doi:10.1111/cdev.13406

Get difference between two smooths

Description

Get difference between two smooths

Usage

get_difference(  model,  series,  compare,  values = NULL,  exclude_terms = NULL,  length_out = 25,  ci_z = 1.96)

Arguments

model

Agam orbam model object.

series

A string specifying the variable that corresponds to the seriesto be plotted on the $x$-axis. If a string is given, the other numericvariables in the model are set to their mean value, unless specific valuesare given invalues. If a character vector of two strings is given, thetwo variables will be taken as the elements of a tensor product smooth.This allows the user to plot 2D raster plots.

compare

A named list of factor levels to compare.

values

User supplied values for specific variables as a named list.

exclude_terms

Terms to be excluded from the prediction. Term namesshould be given as they appear in the model summary (for example,"s(x0,x1)").

length_out

An integer indicating how many values to use along thenumeric variables for predicting the response (the default is10).

ci_z

The z-value for calculating the CIs (the default is1.96 for95 percent CI).

Value

A tibble with the difference smooth.

Examples

library(mgcv)set.seed(10)data <- gamSim(4)model <- gam(y ~ s(x2, by = fac) + s(x0), data = data)get_difference(model, "x2", list(fac = c("1", "2")))

Plot methods for tidygam objects

Description

Plotting methods fortidygam objects.

Usage

## S3 method for class 'tidygam'plot(x, series = NULL, comparison = NULL, raster_interp = FALSE, ...)

Arguments

x

Atidygam object (seepredict_gam()).

series

comparison

Name of a categorical predictor to compare as a string.

raster_interp

Whether to linearly interpolate when plotting a tensorproduct smooth/interaction. It makes sense only whenseries has twovariables. The default isFALSE.

...

Arguments passed toplot().

Value

Aggplot object.

Examples

library(mgcv)set.seed(10)sim_data <- gamSim(4)model_1 <- gam(y ~ s(x2, by = fac) + s(x0), data = sim_data)preds_1 <- predict_gam(model_1, length_out = 50, exclude_terms = "s(x0)")plot(preds_1, "x2")preds_2 <- predict_gam(model_1, length_out = 100, values = list(x0 = 0))plot(preds_2, "x2", "fac")library(ggplot2)plot(preds_2, "x2", "fac") +  scale_fill_brewer(type = "qual") +  scale_color_brewer(type = "qual")# Plotting tensor product smooths/interactionsmodel_2 <- gam(y ~ te(x0, x2, by = fac), data = sim_data)preds_3 <- predict_gam(model_2)preds_3 %>% plot(series = c("x0", "x2"), comparison = "fac")

Plot methods for tidygam.diff objects

Description

Plotting methods fortidygam.diff objects.

Usage

## S3 method for class 'tidygam.diff'plot(x, ..., sig = TRUE, sig_col = "red", sig_alpha = 0.25)

Arguments

x

Atidygam.diff object (seeget_difference()).

...

Arguments passed toplot().

sig

Shade the interval(s) where the difference smooth does not include 0 (default isTRUE).

sig_col

Colour for the shading (default is"red").

sig_alpha

Alpha level for the shading (default is0.25)

Value

Aggplot object.

Examples

library(mgcv)set.seed(10)data <- gamSim(4)model <- gam(y ~ s(x2, by = fac) + s(x0), data = data)model_diff <- get_difference(model, "x2", list(fac = c("1", "2")))plot(model_diff)

Get predictions from a GAM model

Description

Return predictions from a GAM model generated with mgcv. The output can be plotted withplot().

Usage

predict_gam(  model,  length_out = 10,  values = NULL,  series = NULL,  exclude_terms = NULL,  ci_z = 1.96,  tran_fun = NULL,  separate = NULL,  sep_by = "\\.")

Arguments

model

Agam orbam model object.

length_out

An integer indicating how many values to use along thenumeric variables for predicting the response (the default is10).

values

User supplied values for specific variables as a named list.

series

exclude_terms

Terms to be excluded from the prediction. Term namesshould be given as they appear in the model summary (for example,"s(x0,x1)").

ci_z

The z-value for calculating the CIs (the default is1.96 for95 percent CI).

tran_fun

Function to use for transforming the predicted values and CIs.

separate

Names list of factor interaction variables to be separated.

sep_by

Character to separate by (the default is⁠\\.⁠).

Value

A tibble with predictions.

Examples

library(mgcv)set.seed(10)sim_data_1 <- gamSim(1, n = 200, scale = 2)model <- gam(y ~ x0 + s(I(x1^2)) + s(x2) + offset(x3), data = sim_data_1)predict_gam(model)predict_gam(model, values = list(x0 = mean(sim_data_1$x0)))predict_gam(model, series = "x2")predict_gam(model, exclude_terms = "s(I(x1^2))")# By-variablessim_data_2 <- gamSim(4)model_2 <- gam(y ~ s(x2, by = fac) + s(x0), data = sim_data_2)predict_gam(model_2)# Poisson datasim_data_3 <- sim_data_2sim_data_3$y <- round(sim_data_2$y) + 20model_3 <- gam(y ~ s(x2, by = fac), data = sim_data_3, family = poisson)predict_gam(model_3, length_out = 50)predict_gam(model_3, length_out = 50, tran_fun = exp)# Bivariate smoothsmodel_4 <- gam(y ~ te(x1, x2), data = sim_data_1)predict_gam(model_4)

ERP to structural violation in music and language

Description

This data table contains ERP amplitude data from 39 subjects listening to speech and music.

Usage

struct

Format

A tibble with 17160 observations and 6 variables:

t: Time from stimulus onset in milliseconds.
electrode: Electrode number.
voltage: Electrode voltage at time t.
stimulus.condition: Language vs music.
grammar.condition: Structural type (grammatical vs ungrammatical).

Source

doi:10.31234/osf.io/e9w3v