Description Usage Arguments Value Author(s) See Also Examples
Obtain gradients of an ordered or unordered categorical (‘factor’) variable in nonparametric kernel regression models or categorical regression splines.
1 2 3 | npregGradFactor( x, varName, onlyOwnLevels = TRUE )
crsGradFactor( x, varName, onlyOwnLevels = TRUE )
|
x |
an object of class |
varName |
name of the categorical variable. |
onlyOwnLevels |
logical. If |
A matrix that contain the effects of each level of the chosen factor variable (except for the first level) on the dependent variable. In case of an unordered categorical variable, the current level is compared to the first level. In case of an ordered categorical variable, the current level is compared to the previous level. Each row of the matrix corresponds to an observation, while each column corresponds to a level.
Arne Henningsen
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 | data( "oecdpanel" )
library( "miscTools" )
############## unordered factor ###################
oecdpanel$yearFactor <- factor( oecdpanel$year )
npModel <- npreg( growth ~ yearFactor + initgdp,
regtype = "ll", gradients = TRUE, data = oecdpanel )
round( gradients( npModel ), 3 )
yearGrad <- npregGradFactor( npModel, "yearFactor" )
round( yearGrad, 3 )
colMeans( yearGrad, na.rm = TRUE )
colMedians( yearGrad, na.rm = TRUE )
all.equal( gradients( npModel )[ , 1 ], rowSums( yearGrad, na.rm = TRUE ) )
yearGradAll <- npregGradFactor( npModel, "yearFactor", onlyOwnLevels = FALSE )
round( yearGradAll, 3 )
colMeans( yearGradAll )
colMedians( yearGradAll )
all.equal( yearGradAll[ !is.na( yearGrad ) ], yearGrad[ !is.na( yearGrad ) ] )
############## ordered factor ###################
oecdpanel$yearOrdered <- ordered( oecdpanel$year )
npModelOrdered <- npreg( growth ~ yearOrdered + initgdp,
regtype = "ll", gradients = TRUE, data = oecdpanel )
round( gradients( npModelOrdered ), 3 )
yearGradOrdered <- npregGradFactor( npModelOrdered, "yearOrdered" )
round( yearGradOrdered, 3 )
colMeans( yearGradOrdered, na.rm = TRUE )
colMedians( yearGrad, na.rm = TRUE )
all.equal( gradients( npModelOrdered )[ oecdpanel$year != 1965, 1 ],
rowSums( yearGradOrdered, na.rm = TRUE )[ oecdpanel$year != 1965 ] )
yearGradOrderedAll <- npregGradFactor( npModelOrdered, "yearOrdered",
onlyOwnLevels = FALSE )
round( yearGradOrderedAll, 3 )
colMeans( yearGradOrderedAll )
colMedians( yearGradOrderedAll )
all.equal( yearGradOrderedAll[ !is.na( yearGradOrdered ) ],
yearGradOrdered[ !is.na( yearGradOrdered ) ] )
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.