toolRegression: toolRegression

In pik-piam/regressionworlddata: Regression analysis based on global datasets

toolRegression

Description

Regression model for the correlation of a denominator and quotient to the GDP, allowing for an additional driver z next to income.

Usage

toolRegression(
  denominator,
  quotient = NULL,
  func = y ~ (a * x)/(b + x),
  x = "IHME_USD05_PPP_pc",
  z = NULL,
  ylab = NULL,
  xlab = NULL,
  data = NULL,
  countries_nlsAddLines = NULL,
  weight = "pop",
  x_log10 = FALSE,
  crossvalid_sample = NULL,
  crossvalid_drawing = 1,
  ...
)

Arguments

denominator	denominator of the dependent variable that shall be estimated using the regression
quotient	quotient of the dependent variable that shall be estimated using the regression
func	functional relation for the regression, shall be in the format y~f(x,...) with x being gdp, y being denominator/quotient, and f() being any type of functional relationship. ... can inlcude either z or parameters to be estimated.
x	independet variable, by default income
z	additional independet variable
ylab	name of y axis
xlab	name of x axis
data	data can be provided if Data shall not be derived by mrcommons:::calcCollectFoodDemandRegressionData()
countries_nlsAddLines	the number of weightiest countries or the name of countries that shall be plotted by lines in the plot
weight	the weight
x_log10	passed on to nlsregression()
crossvalid_sample	sample name from madrat used for crossvalidation. Name is built as follows: crossvalid_seedX_kY X is the random seed, Y is the number of drawings. The combination of all drawings is the full sample.
crossvalid_drawing	selected drawing of k in crossvalidsample
...	further attributes that will be handed on to nlsregression(): An additional explanatory variable z can be added. A regression model has to be chosen. Startvalues can be predetermained.

Value

regression plot and the parameters from nlsregression

Author(s)

Antonia Walther, Benjamin Leon Bodirsky

See Also

calcOutput

Examples

## Not run:  

toolRegression(denominator=livestock,
                     func=y~(a*x)/(b+x),
                     z=NULL,
                     startvalues=list(a=1100,b=7770)
                     )
                     
toolRegression(denominator=findset("kap"),
                     quotient=findset("kfo"),
                     func=y~(a*x)/(b+x),
                     z=NULL,
                     startvalues=list(a=0.5,b=7770)
                     )
                     

## End(Not run)

pik-piam/regressionworlddata documentation built on June 6, 2024, 11:08 p.m.