Description Usage Arguments Value
View source: R/basicFunctionsKernel.R
FlexCoDE Fit Conditional Density Estimation via Regression
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
zTrain |
Responses z used to train the model (matrix with one column; one observation per row) |
zValidation |
Responses z used to tune the model (matrix with one column; one observation per row) |
zTest |
Responses z used to estimate risk of final model (matrix with one column; one observation per row). Default is NULL |
nIMax |
Maximum possible number of components of the series expansion (that is, the function will find the best I<nIMax). Default is 100 |
regressionFunction |
a function indicating which regression method will be used to estimate the expansion coefficients. Currently can be one of |
regressionFunction.extra |
extra parameters to be sent to regression function; see the regression you want to use to check what are the available parameters |
system |
Basis for z. Current options are "Fourier", "Cosine" and "discrete". Default is "Fourier" |
deltaGrid |
Grid of threshold deltas (betwen 0 and 0.5). Default value is seq(0,0.4,0.05). |
chooseDelta |
Should delta, the cutoff to remove spurious bumps, be chosen? |
verbose |
Should we print what we are doing? Default is TRUE. |
xTrain |
Covariates x used to train the model (one observation per row) |
xValidation |
Covariates x used to tune the model (one observation per row; same number of columns as xTrain) |
xTest |
Covariates x used to estimate risk of final model (one observation per row; same number of columns as xTrain). Default is NULL |
Returns the fitted estimated conditional density, and object of the class FlexCoDE. The return value is an object with the following components:
zMin, zMax |
Minimum and maximum value of z |
nIMax |
Maximum number of expansion coefficients (user input). Default is minimum between 25 and number of training samples. |
system |
Basis used for expanding the response |
zTrain |
zTrain (user input) |
xTrain |
xTrain (user input) |
regressionObject |
Object with fitted regressions. Class and content depend on which regression method was chosen by user |
errors |
Estimated errors for each value of I (number of expansion coefficients) using validation set |
bestI |
Optimal number of I according to validation set |
bestError |
Estimated error of model with bestI expansion terms according to validation set |
bestDelta |
Optimal value of threshold delta according to validation set |
estimatedRisk |
(If user provides xTest and zTest) Estimated risk (error) according to test set) |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.