se: Sum of Squares of Error Statistics
| se | R Documentation |
Sum of Squares of Error Statistics
Description
Computes the residual sum of squares and the root of the mean of for squares
for models: nls, nls.lm, and
models CDFmodel and CDFmodelList obatained with [fitCDF].
Usage
se(model, stat)
## S3 method for class 'nls'
se(model, stat = c("rmse", "sme"))
## S3 method for class 'nls.lm'
se(model, stat = c("rmse", "sme"))
## S3 method for class 'CDFmodel'
se(model, stat = c("rmse", "sme"))
## S3 method for class 'CDFmodelList'
se(model, stat = c("rmse", "sme"))
Arguments
model |
An object from one of the classes: |
stat |
Select which statistic to return: 1) The root mean of the square of error or the square errors of residuals. |
Value
The requested statistic for each model.
Examples
## Simulate data from Normal distribution
## and search for the best fitted model
x <- rnorm(1000, mean = 5, sd = 2)
models <- fitCDF2(x, distNames = c(1, 7, 11, 12))
## Bases on the AIC criteria the best model is Weibull 3P
models
## In the current case Weibull is also the model with the lowest error
se(model)
