https://github.com/mayer79/actools

Tools for actuary science
https://github.com/mayer79/actools

Last synced: 1 day ago
JSON representation

Tools for actuary science

• Host: GitHub
• URL: https://github.com/mayer79/actools
• Owner: mayer79
• License: gpl-3.0
• Created: 2019-06-02T10:19:20.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-06-04T06:46:21.000Z (over 5 years ago)
• Last Synced: 2024-10-04T12:56:53.509Z (2 months ago)
• Language: R
• Size: 273 KB
• Stars: 1
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Changelog: NEWS.md
  • License: LICENSE

Awesome Lists containing this project

• jimsghstars - mayer79/actools - Tools for actuary science (R)

README

        
# actools

Tools for actuarial science.

 

## Description

The package can e.g. be used to extract coefficient tables from an all-factor linear regression model with arbitrarily interaction levels fitted by `lm`. 

## Examples

``` r

# devtools::install_github("mayer79/actools/Release/actools")

library(actools)

CO2[["conc"]] <- factor(CO2[["conc"]])

fit <- lm(uptake ~ Type:Treatment + conc:Treatment + Type:conc, data = CO2)

(coefT <- coefTables(fit, rebase = TRUE))

head(pred_classic <- predict(fit, CO2))

head(pred_fancy <- predict(coefT, CO2))

    

perf(CO2$uptake, pred_classic, lab = "classic")

perf(CO2$uptake, pred_fancy, lab = "fancy")  

# Same with logs

fit <- lm(log(uptake) ~ Type:Treatment + conc:Treatment + Type:conc, data = CO2)

(coefT <- coefTables(fit, rebase = TRUE))

head(pred_classic <- exp(predict(fit, CO2)))

head(pred_fancy <- predict(coefT, CO2, trafo = exp))

perf(CO2$uptake, pred_classic, lab = "classic")

perf(CO2$uptake, pred_fancy, lab = "fancy")

```

## References