datana (version 1.1.2)

valestamod: Function to create a table with the prediction statistics by model.

Description

Creates a table with the prediction statistics for previously fitted models, based on the observed data.

Usage

valestamod(
  data = data,
  y.obs = "y.obs",
  y.pred = "y.pred",
  model = "model",
  want.by.valcl = FALSE,
  val.class = NA,
  want.all.outputs = FALSE
)

Value

The main output is a table having as number of rows the total number of fitted models, and number of columns the statistics being computed. By default the statistics implemented in the valesta() function are computed.

Arguments

data

a dataframe having the predicted and observed values of the response variable for a set of models.

y.obs

a character giving the column name of the response variable for the data.

y.pred

a character giving the column name of the predicted value for the response variable giving the predictor(s) variable(s) values for the data and the respective fitted model.

model

a character giving the column name for the name of previously fitted model(s).

want.by.valcl

A logical option for requesting to also computed the prediction statistics by validation classes, which are stored in the column defined in val.class. By default is set to FALSE.

val.class

If validation classes were assigned to each observation, this option corresponds to character giving the column name of the validation class. By default this option is set to NA, meaning is not available.

want.all.outputs

A logical option to save a full set of result elements in the output, thus the output is class list. By default is set to FALSE.

Author

Christian Salas-Eljatib and Marcos Marivil.

Details

The function computes prediction statistics for a previously fitted model, and prepare an output summarizing the results to facilitate the comparison among models.

References

• Salas C. 2002. Ajuste y validación de ecuaciones de volumen para un relicto del bosque de roble-laurel-lingue. Bosque 23(2):81–92. tools:::Rd_expr_doi("10.4067/S0717-92002002000200009").

Examples

#Creates a fake dataframe
set.seed(1234);
Y=rnorm(30, 30,9);X=rnorm(30, 450,133); Z=rbeta(30, .1,2)
df <- as.data.frame(cbind(Y, X,Z))
## Fitting some models
mod1 <- lm(Y~X, data=df)
mod2 <- lm(Y~X+I(X^2), data=df)
mod3 <- lm(Y~Z+I(X^2), data=df)
## Preparing the format of the input-data for the function
df.m1<-df.m2<-df.m3<-df
df.m1$model<-"mod1";df.m1$y.aju=fitted(mod1)
df.m2$model<-"mod2";df.m2$y.aju=fitted(mod2)
df.m3$model<-"mod3";df.m3$y.aju=fitted(mod3)
dfypredmod<-rbind(df.m1,df.m2,df.m3)
head(dfypredmod)
table(dfypredmod$model)
# Example
valestamod(data=dfypredmod,y.obs="Y",y.pred="y.aju")