| Type: | Package |
| Title: | Interactive Model Exploration using 'GGobi' |
| Version: | 0.3.1 |
| Author: | Hadley Wickham <h.wickham@gmail.com> |
| Maintainer: | Hadley Wickham <h.wickham@gmail.com> |
| Description: | Exploratory model analysis with http://ggobi.org. Fit and graphical explore ensembles of linear models. |
| License: | MIT + file LICENSE |
| URL: | https://github.com/hadley/meifly |
| BugReports: | https://github.com/hadley/meifly/issues |
| Imports: | leaps, MASS, plyr |
| Encoding: | UTF-8 |
| RoxygenNote: | 7.2.0 |
| NeedsCompilation: | no |
| Packaged: | 2022-05-20 16:23:54 UTC; hadleywickham |
| Repository: | CRAN |
| Date/Publication: | 2022-05-20 16:40:02 UTC |
Calculcate coefficients for all models in ensemble. Returns raw, t-value, absolute t-value, and standardised coefficent values.
Description
Calculcate coefficients for all models in ensemble. Returns raw, t-value, absolute t-value, and standardised coefficent values.
Usage
## S3 method for class 'ensemble'
coef(object, ...)
Arguments
object |
ensemble of models |
... |
other arguments ignored |
General ensemble of models from models in global workspace'
Description
General ensemble of models from models in global workspace'
Usage
findmodels(modeltype = "lm", dataset, pattern)
Arguments
modeltype |
model class |
dataset |
if specified, all models must use this dataset |
pattern |
pattern of model object names to match |
Fit all combinations of x variables ($2^p$).
Description
This technique generalises fitbest. While it is much
slower it will work for any type of model.
Usage
fitall(y, x, method = "lm", ...)
Arguments
y |
vector y values |
x |
matrix of x values |
method |
|
... |
other arguments passed on to |
Examples
y <- swiss$Fertility
x <- swiss[, -1]
mods <- fitall(y, x, "lm")
Use the leaps package to generate the best subsets.
Description
Use the leaps package to generate the best subsets.
Usage
fitbest(formula, data, nbest = 10, ...)
Arguments
formula |
model formula |
data |
data frame |
nbest |
number of subsets of each size to record |
... |
other arguments passed to |
Examples
y <- swiss$Fertility
mods <- fitbest(Fertility ~ ., swiss)
Generate linear models by bootstrapping observations
Description
Generate linear models by bootstrapping observations
Usage
lmboot(formula, data, n = 100)
Arguments
formula |
model formula |
data |
data set |
n |
number of bootstrapped data sets to generate |
Interactive model ensemble exploration.
Description
Interactive model ensemble exploration.
Create a new ensemble of models.
Description
Create a new ensemble of models.
Usage
new_ensemble(models, data)
Arguments
models |
list of models |
data |
associated data frame |
Calculate residuals for all models in ensemble.
Description
Calculate residuals for all models in ensemble.
Usage
## S3 method for class 'ensemble'
residuals(object, ...)
Arguments
object |
ensemble of models |
... |
other arguments ignored |
Value
data.frame of class resid_ensemble
See Also
Returns degrees of freedom, log likelihood, R-squared, AIC, BIC and adjusted R-squared.
Description
Returns degrees of freedom, log likelihood, R-squared, AIC, BIC and adjusted R-squared.
Usage
## S3 method for class 'ensemble'
summary(object, ...)
Arguments
object |
ensemble of models |
... |
other arguments ignored |
Summarise residuals from ensemble.
Description
Summarise residuals from ensemble.
Usage
## S3 method for class 'resid_ensemble'
summary(object, data = attr(object, "data"), ...)
Arguments
object |
model residuals from |
data |
associated data set |
... |
other arguments ignored |
Summarise variable ensemble.
Description
Provides variable level statistics.
Usage
## S3 method for class 'variable_ensemble'
summary(object, ...)
Arguments
object |
ensemble of models |
... |
other arguments ignored |