Package 'regressoR'

Title: Regression Data Analysis System
Description: Perform a supervised data analysis on a database through a 'shiny' graphical interface. It includes methods such as linear regression, penalized regression, k-nearest neighbors, decision trees, ada boosting, extreme gradient boosting, random forest, neural networks, deep learning and support vector machines.
Authors: Oldemar Rodriguez [aut, cre], Andres Navarro D. [ctb, prg], Diego Jimenez A. [ctb, prg], Ariel Arroyo S. [ctb, prg], Joseline Quiros M. [ctb, prg]
Maintainer: Oldemar Rodriguez <[email protected]>
License: GPL (>=2)
Version: 3.0.2
Built: 2024-10-31 20:28:01 UTC
Source: https://github.com/PROMiDAT/regressoR

Help Index

The application server-side

Description

The application server-side

Usage

app_server(input, output, session)

Arguments

input, output, session

Internal parameters for shiny. DO NOT REMOVE.

as_string_c

Description

creates a string representative of a vector

Usage

as_string_c(vect, quote = TRUE)

Arguments

vect

a vector with values
quote

a logical value. If TRUE, the values on the vector will be surrounded by quotes.

Examples

as_string_c(c("A", "B", "C"))
as_string_c(c(5, 6, 7))
as_string_c(c(5, 6, 7), quote = FALSE)
as_string_c(iris$Species)

boosting_importance_plot

Description

generates the graph of variable importance.

Usage

boosting_importance_plot(
  model,
  titles = c("Importancia de Variables segun Influencia Relativa",
    "Influencia Relativa", "Variable")
)

Arguments

model

boosting model(gbm).
titles

Labels on the chart

calibrate_boosting

Description

helps to get the maximum of n.minobsinnode and bag.fraction values with which no error is generated in the model.

Usage

calibrate_boosting(data)

Arguments

data

the name of the learning data.

See Also

gbm

Examples

calibrate_boosting(iris)

coef_lambda

Description

get penalized regression coefficients.

Usage

coef_lambda(data, variable.pred, model, log.lambda = NULL)

Arguments

data

dataframe
variable.pred

the name of the variable to be predicted.
model

a penalized regression model(cv.glmnet).
log.lambda

numerical. Logarithm of lambda in case you don't want to use the optimal lambda.

Create disjunctive columns to a data.frame.

Description

Create disjunctive columns to a data.frame.

Usage

datos.disyuntivos(data, var)

Arguments

data

a data.frame object.
var

the column name to apply disjunctive code.

Value

data.frame

Author(s)

Diego Jimenez <[email protected]>

Examples

datos.disyuntivos(iris, "Species")

disp_models

Description

this function generates the call code of the scatter function.

Usage

disp_models(prediction, model_name, var_pred)

Arguments

prediction

the name of the prediction object.
model_name

the name of the model.
var_pred

the name of the variable to be predicted.

Examples

disp_models("prediction.knn", "KNN", "Species")

dt_plot

Description

makes the graph of the tree.

Usage

dt_plot(model)

Arguments

model

a decision trees model(rpart).

e_coeff_landa

Description

Graph the coefficients and lambdas of a cv.glmnet model

Usage

e_coeff_landa(
  cv.glm,
  log.lambda = NULL,
  titles = c("Coeficientes", "Seleccionado", "Automatico")
)

Arguments

cv.glm

a cv.glmnet model.
log.lambda

number that specifies the logarithm of the selected lambda
titles

labels on the chart

Value

echarts4r plot

Author(s)

Ariel Arroyo <[email protected]>

See Also

cv.glmnet

Eval character vectors to JS code

Description

Eval character vectors to JS code

Usage

e_JS(...)

Arguments

...

character vectors to evaluate

Author(s)

Joseline Quiros <[email protected]>

Examples

e_JS('5 * 3')

e_posib_lambda

Description

Graph a cv.glmnet model

Usage

e_posib_lambda(
  cv.glm,
  log.lambda = NULL,
  titles = c("Error Cuadratico Medio", "Curva Inferior", "Curva Superior",
    "Seleccionado", "Automatico", "Coeficientes Distintos de Cero")
)

Arguments

cv.glm

a cv.glmnet model.
log.lambda

number that specifies the logarithm of the selected lambda
titles

labels on the chart

Value

echarts4r plot

Author(s)

Ariel Arroyo <[email protected]>

See Also

cv.glmnet

exe

Description

concat and execute a text in R.

Usage

exe(..., envir = parent.frame())

Arguments

...

one or more texts to be concatenated and executed.
envir

the environment in which expr is to be evaluated.

Value

the result of the execute.

Examples

exe("5+5")
exe("5","+","5")
exe("plot(iris$Species)")

extract_code

Description

gets the code of a function in text form.

Usage

extract_code(funcion, envir = parent.frame())

Arguments

funcion

the name of the function to be extracted.
envir

the environment in which expr is to be evaluated.

Examples

extract_code("cat")
extract_code("plot")

parse(text = extract_code("plot"))

general_indices

Description

calculates indices to measure accuracy of a model.

Usage

general_indices(real, prediccion)

Arguments

real

the real values in traning-testing.
prediccion

the prediction values in traning-testing.

Value

a list with the Correlation, Relative Error, Mean Absolute Error and Root Mean Square Error.

Examples

real <- rnorm(45)
prediction <- rnorm(45)
model <- "KNN"
general_indices(real, prediction)

importance_plot_rf

Description

graphs the importance of variables for the random forest model according to the percentage increase in mean square error.

Usage

importance_plot_rf(
  model.rf,
  titles = c("Importancia de Variables Segun el Porcentaje de Incremento del MSE",
    "Aumento porcentual del error cuadratico medio", "Variable")
)

Arguments

model.rf

a random forest model.
titles

labels on the chart

Value

echarts4r plot

Author(s)

Ariel Arroyo <[email protected]>

See Also

randomForest

nn_plot

Description

graph of the neural network.

Usage

nn_plot(model)

Arguments

model

a neural network model(neuralnet)

pairs_power

Description

Generate a pair chart

Usage

pairs_power(data, decimals = 2)

Arguments

data

A DataFrame
decimals

Number of numbers after the decimal point.

plot_pred_rd

Description

graph of variance explained in the predictors according to components used.

Usage

plot_pred_rd(
  model,
  n.comp,
  titles = c("Varianza Explicada en Predictores", "Numero de Componentes",
    "Porcentaje de Varianza Explicada")
)

Arguments

model

a dimension reduction model.
n.comp

the optimum number of components.
titles

labels on the chart

Value

echarts4r plot

Author(s)

Ariel Arroyo <[email protected]>

plot_real_prediction

Description

scatter plot between the actual value of the variable to be predicted and the prediction of the model.

Usage

plot_real_prediction(
  real,
  prediction,
  model = "",
  titles = c("Predicciones vs Valores Reales", "Valor Real", "Prediccion")
)

Arguments

real

the real values in traning-testing.
prediction

the prediction values in traning-testing.
model

the name of the model of the scatter plot.
titles

Labels on the chart

Value

echarts4r plot

Author(s)

Ariel Arroyo <[email protected]>

plot_RMSE

Description

graph the root mean square error of cross validation according to components used.

Usage

plot_RMSE(
  model,
  n.comp,
  titles = c("RMSE Segun Numero de Componentes", "Numero de Componente", "RMSE")
)

Arguments

model

a dimension reduction model.
n.comp

the optimum number of components.
titles

labels on the chart

Value

echarts4r plot

Author(s)

Ariel Arroyo <[email protected]>

plot_var_pred_rd

Description

graph of the variance explained in the variable to predict according to the components used.

Usage

plot_var_pred_rd(
  model,
  n.comp,
  titles = c("Varianza Explicada en Variable a Predecir", "Numero de Componente",
    "Porcentaje de Varianza Explicada")
)

Arguments

model

a dimension reduction model.
n.comp

the optimum number of components.
titles

labels on the chart

Value

echarts4r plot

Author(s)

Ariel Arroyo <[email protected]>

rd_model

Description

generates a dimension reduction model.

Usage

rd_model(data, variable.pred, mode = 0, scale = TRUE)

Arguments

data

dataframe
variable.pred

the name of the variable to be predicted.
mode

the method of dimension reduction is defined as mode=1 is the MCP, and mode=0 the ACP.
scale

the scale parameter of the model.

See Also

pcr, plsr

rd_prediction

Description

generates the prediction of a dimension reduction model.

Usage

rd_prediction(model, test.data, ncomp = NULL)

Arguments

model

dimension reduction model(pcr/plsr).
test.data

dataframe.
ncomp

a numerical value in case you don't want to use the optimum number of components.

rd_type

Description

returns the name of the method of dimension reduction.

Usage

rd_type(mode.rd = 0)

Arguments

mode.rd

the method of dimension reduction is defined as mode=1 is the MCP, and mode=0 the ACP.

See Also

pcr, plsr

Examples

rd_type(1)
rd_type(0)

rl_coeff

Description

get the information of the coefficients of the linear regression model

Usage

rl_coeff(modelo)

Arguments

modelo

linear regression model

rlr_model

Description

generates a penalized regression model.

Usage

rlr_model(data, variable.pred, alpha = 0, standardize = TRUE)

Arguments

data

dataframe
variable.pred

the name of the variable to be predicted.
alpha

the alpha parameter of the model.
standardize

the standardize parameter of the model.

See Also

glmnet, cv.glmnet

rlr_prediction

Description

generates the prediction of the penalized regression model.

Usage

rlr_prediction(model, test.data, variable.pred, log.lambda = NULL)

Arguments

model

a penalized regression model(cv.glmnet).
test.data

dataframe.
variable.pred

the name of the variable to be predicted.
log.lambda

numerical. Logarithm of lambda in case you don't want to use the optimal lambda.

rlr_type

Description

returns the name of the penalty according to the alpha.

Usage

rlr_type(alpha_rlr = 0)

Arguments

alpha_rlr

the penalty is defined as alpha=1 is the lasso penalty, and alpha=0 the ridge penalty.

See Also

glmnet

Examples

rlr_type(1)
rlr_type(0)

Run the Shiny Application

Description

Run the Shiny Application

Usage

run_app(...)

Arguments

...

A series of options to be used inside the app.

summary_indices

Description

summarizes a variable by returning the minimum, first quartile, third quartile and maximum value.

Usage

summary_indices(data)

Arguments

data

a numeric vector.

Examples

summary_indices(iris$Sepal.Length)