| Title: | Delete or Merge Regressors Algorithms for Linear and Logistic Model Selection and High-Dimensional Data |
|---|---|
| Description: | Model selection algorithms for regression and classification, where the predictors can be continuous or categorical and the number of regressors may exceed the number of observations. The selected model consists of a subset of numerical regressors and partitions of levels of factors. Szymon Nowakowski, Piotr Pokarowski, Wojciech Rejchel and Agnieszka Sołtys, 2023. Improving Group Lasso for High-Dimensional Categorical Data. In: Computational Science – ICCS 2023. Lecture Notes in Computer Science, vol 14074, p. 455-470. Springer, Cham. <doi:10.1007/978-3-031-36021-3_47>. Aleksandra Maj-Kańska, Piotr Pokarowski and Agnieszka Prochenka, 2015. Delete or merge regressors for linear model selection. Electronic Journal of Statistics 9(2): 1749-1778. <doi:10.1214/15-EJS1050>. Piotr Pokarowski and Jan Mielniczuk, 2015. Combined l1 and greedy l0 penalized least squares for linear model selection. Journal of Machine Learning Research 16(29): 961-992. <https://www.jmlr.org/papers/volume16/pokarowski15a/pokarowski15a.pdf>. Piotr Pokarowski, Wojciech Rejchel, Agnieszka Sołtys, Michał Frej and Jan Mielniczuk, 2022. Improving Lasso for model selection and prediction. Scandinavian Journal of Statistics, 49(2): 831–863. <doi:10.1111/sjos.12546>. |
| Authors: | Agnieszka Prochenka-Sołtys [aut] (previous maintainer for versions <=
0.2.0),
Piotr Pokarowski [aut],
Szymon Nowakowski [aut, cre] |
| Maintainer: | Szymon Nowakowski <[email protected]> |
| License: | GPL-2 |
| Version: | 0.4.0.9000 |
| Built: | 2025-02-17 02:43:52 UTC |
| Source: | https://github.com/szymonnowakowski/dmrnet |
Model selection algorithms for regression and classification, where the predictors can be continuous or categorical and the number of regressors may exceed the number of observations. The selected model consists of a subset of numerical regressors and partitions of levels of factors.
Similar in use to glmnet. It consists of the following functions:
DMR - Model selection algorithm for p<n; produces a path of models.
DMRnet - Model selection algorithm both for p<n and for p>=n; produces a path of models.
print.DMR, coef.DMR, plot.DMR, predict.DMR - Functions for inspection of the models on the path.
gic.DMR, cv.DMR, cv.DMRnet - Functions for final model selection, resulting with one model from the path.
coef.gic.DMR, coef.cv.DMR, plot.gic.DMR, plot.cv.DMR, predict.gic.DMR, predict.cv.DMR - Functions for inspection of the final model.
miete, promoter - Two data sets used for vignettes, examples, etc.
For more information see a friendly "Getting started" vignette:
Agnieszka Prochenka-Sołtys, Piotr Pokarowski, Szymon Nowakowski
Maintainer: Szymon Nowakowski <[email protected]>
Aleksandra Maj-Kańska, Piotr Pokarowski and Agnieszka Prochenka, 2015. Delete or merge regressors for linear model selection. Electronic Journal of Statistics 9(2): 1749-1778. doi:10.1214/15-EJS1050
Piotr Pokarowski and Jan Mielniczuk, 2015. Combined l1 and greedy l0 penalized least squares for linear model selection. Journal of Machine Learning Research 16(29): 961-992. https://www.jmlr.org/papers/volume16/pokarowski15a/pokarowski15a.pdf
Piotr Pokarowski, Wojciech Rejchel, Agnieszka Sołtys, Michał Frej and Jan Mielniczuk, 2022. Improving Lasso for model selection and prediction. Scandinavian Journal of Statistics, 49(2): 831–863. doi:10.1111/sjos.12546
## Not run: vignette("getting-started", package="DMRnet") ## End(Not run)## Not run: vignette("getting-started", package="DMRnet") ## End(Not run)
Extracts coefficients from a cv.DMR object (for the model with minimal cross-validated error /the default/ or the smallest model falling under the upper curve of a prediction error plus one standard deviation).
## S3 method for class 'cv.DMR' coef(object, md = "df.min", ...)## S3 method for class 'cv.DMR' coef(object, md = "df.min", ...)
object |
Fitted |
md |
Value of the model dimension parameter at which predictions are required. The default is |
... |
Further arguments passed to or from other methods. |
Similar to other coef methods, this function extracts coefficients from a fitted cv.DMR object.
Vector of coefficients.
## cv.DMR for linear regression set.seed(13) data(miete) y <- miete$rent X <- miete$area cv = cv.DMR(X,y) coef(cv)## cv.DMR for linear regression set.seed(13) data(miete) y <- miete$rent X <- miete$area cv = cv.DMR(X,y) coef(cv)
Extracts coefficients from a DMR object.
## S3 method for class 'DMR' coef(object, df = NULL, ...)## S3 method for class 'DMR' coef(object, df = NULL, ...)
object |
Fitted |
df |
Number of parameters in the model for which coefficients are required. Default is the entire path of models. |
... |
Further arguments passed to or from other methods. |
Similar to other coef methods, this function extracts coefficients from a fitted DMR object.
Vector or matrix of coefficients.
data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) coef(m, df = 12)data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) coef(m, df = 12)
Extracts coefficients from a gic.DMR object (for the model with minimal gic).
## S3 method for class 'gic.DMR' coef(object, ...)## S3 method for class 'gic.DMR' coef(object, ...)
object |
Fitted |
... |
Further arguments passed to or from other methods. |
Similar to other coef methods, this function extracts coefficients from a fitted gic.DMR object for the model with minimal gic.
Vector of coefficients.
data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) g <- gic.DMR(m, c = 2.5) coef(g)data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) g <- gic.DMR(m, c = 2.5) coef(g)
Executes k-fold cross-validation for DMR and returns a value for df.
cv.DMR( X, y, family = "gaussian", clust.method = "complete", lam = 10^(-7), nfolds = 10, indexation.mode = "GIC" )cv.DMR( X, y, family = "gaussian", clust.method = "complete", lam = 10^(-7), nfolds = 10, indexation.mode = "GIC" )
X |
Input data frame, of dimension n x p; |
y |
Response variable. Numerical for |
family |
Response type; one of: |
clust.method |
Clustering method used for partitioning levels of factors; see function hclust in package stats for details. |
lam |
The amount of penalization in ridge regression (used for logistic regression in order to allow for parameter estimation in linearly separable setups) or the amount of matrix regularization in case of linear regression. Used only for numerical reasons. The default is 1e-7. |
nfolds |
Number of folds in cross-validation. The default value is 10. |
indexation.mode |
How the cross validation algorithm should index the models for internal quality comparisons; one of: |
cv.DMR algorithm does cross-validation for DMR with nfolds folds. The df for the minimal estimated prediction error is returned.
An object with S3 class "cv.DMR" is returned, which is a list with the ingredients of the cross-validation fit.
df (number of parameters) of the model with minimal cross-validated error.
df (number of parameters) of the smallest model falling under the upper curve of a prediction error plus one standard deviation.
Fitted DMR object for the full data.
The mean cross-validated error for the entire sequence of models.
The fold assignments used.
plot.cv.DMR for plotting, coef.cv.DMR for extracting coefficients and predict.cv.DMR for prediction.
## cv.DMR for linear regression set.seed(13) data(miete) ytr <- miete$rent[1:1500] Xtr <- miete$area[1:1500] Xte <- miete$area[1501:2053] cv <- cv.DMR(Xtr, ytr) print(cv) plot(cv) coef(cv) ypr <- predict(cv, newx = Xte)## cv.DMR for linear regression set.seed(13) data(miete) ytr <- miete$rent[1:1500] Xtr <- miete$area[1:1500] Xte <- miete$area[1501:2053] cv <- cv.DMR(Xtr, ytr) print(cv) plot(cv) coef(cv) ypr <- predict(cv, newx = Xte)
Executes k-fold cross-validation for DMR and returns a value for df.
cv.DMRnet( X, y, family = "gaussian", o = 5, nlambda = 100, lam = 10^(-7), interc = TRUE, maxp = ifelse(family == "gaussian", ceiling(length(y)/2), ceiling(length(y)/4)), nfolds = 10, indexation.mode = "GIC", algorithm = "DMRnet", clust.method = ifelse(algorithm == "glamer", "single", "complete") )cv.DMRnet( X, y, family = "gaussian", o = 5, nlambda = 100, lam = 10^(-7), interc = TRUE, maxp = ifelse(family == "gaussian", ceiling(length(y)/2), ceiling(length(y)/4)), nfolds = 10, indexation.mode = "GIC", algorithm = "DMRnet", clust.method = ifelse(algorithm == "glamer", "single", "complete") )
X |
Input data frame, of dimension n x p; each row is an observation vector. Columns can be numerical or integer for continuous predictors or factors for categorical predictors. |
y |
Response variable. Numerical for |
family |
Response type; one of: |
o |
Parameter of the group lasso screening step, described in |
nlambda |
Parameter of the group lasso screening step, described in |
lam |
The amount of penalization in ridge regression (used for logistic regression in order to allow for parameter estimation in linearly separable setups) or the amount of matrix regularization in case of linear regression. Used only for numerical reasons. The default value is 1e-7. |
interc |
Should intercept(s) be fitted (the default, |
maxp |
Maximal number of parameters of the model, smaller values result in quicker computation. |
nfolds |
Number of folds in cross-validation. The default value is 10. |
indexation.mode |
How the cross validation algorithm should index the models for internal quality comparisons; one of: |
algorithm |
The algorithm to be used; for partition selection (merging levels) use one of: |
clust.method |
Clustering method used for partitioning levels of factors; see function hclust in package stats for details. |
cv.DMRnet algorithm does nfold-fold cross-validation for DMRnet. The df for the minimal estimated prediction error is returned.
An object with S3 class "cv.DMR" is returned, which is a list with the ingredients of the cross-validation fit.
df (number of parameters) of the model with minimal cross-validated error.
df (number of parameters) of the smallest model falling under the upper curve of a prediction error plus one standard deviation.
Fitted DMR object for the full data.
The mean cross-validated error for the entire sequence of models.
The fold assignments used.
plot.cv.DMR for plotting, coef.cv.DMR for extracting coefficients and predict.cv.DMR for prediction.
## cv.DMRnet for linear regression set.seed(13) data(miete) ytr <- miete$rent[1:1500] Xtr <- miete$area[1:1500] Xte <- miete$area[1501:2053] cv <- cv.DMRnet(Xtr, ytr) print(cv) plot(cv) coef(cv) ypr <- predict(cv, newx = Xte)## cv.DMRnet for linear regression set.seed(13) data(miete) ytr <- miete$rent[1:1500] Xtr <- miete$area[1:1500] Xte <- miete$area[1501:2053] cv <- cv.DMRnet(Xtr, ytr) print(cv) plot(cv) coef(cv) ypr <- predict(cv, newx = Xte)
Fits a path of linear (family="gaussian") or logistic (family="binomial") regression models, where the number of parameters changes from 1 to p (p is the number of columns in the model matrix). Models are subsets of continuous predictors and partitions of levels of factors in X.
DMR( X, y, family = "gaussian", clust.method = "complete", lam = 10^(-7), lambda = NULL )DMR( X, y, family = "gaussian", clust.method = "complete", lam = 10^(-7), lambda = NULL )
X |
Input data frame; each row is an observation vector; each column can be numerical or integer for a continuous predictor or a factor for a categorical predictor; DMR works only if p<n (n is the number of observations, p the number of columns in the model matrix), for p>=n see |
y |
Response variable; Numerical for |
family |
Response type; one of: |
clust.method |
Clustering method used for partitioning levels of factors; see function hclust in package stats for details. |
lam |
The amount of penalization in ridge regression (used for logistic regression in order to allow for parameter estimation in linearly separable setups) or the amount of matrix regularization in case of linear regression. Used only for numerical reasons. The default is 1e-7. |
lambda |
The net of lambda values. It is optional and serves only for consistency with |
DMR algorithm is based on a traditional stepwise method.
A nested family of models is built based on the values of squared Wald statistics:
1. For each continuous variable the squared Wald statistic is calculated for a hypothesis that the variable is equal to zero (it should be deleted).
2. For each factor a dissimilarity matrix is constructed using squared Wald statistics for hypotheses that two parameters are equal (the two levels of factor should be merged). Next, hierarchical clustering is preformed using the dissimilarity matrix. All cutting heights are recorded.
3. Squared Wald statistics and cutting heights and values of from steps 2 and 3 are concatenated and sorted, resulting in vector h.
4. Nested family of models of size 1 to p is built by accepting hypotheses according to increasing values in vector h.
An object with S3 class "DMR", which is a list with the ingredients:
beta |
Matrix p times p of estimated parameters; each column corresponds to a model on the nested path having from p to 1 parameter (denoted as df). |
df |
Vector of degrees of freedom; from p to 1. |
rss/loglik |
Measure of fit for the nested models: rss (residual sum of squares) is returned for |
n |
Number of observations. |
levels.listed |
Minimal set of levels of respective factors present in data. |
lambda |
The net of lambda values used in the screening step, empty vector in case of |
arguments |
List of the chosen arguments from the function call. |
interc |
If the intercept was fitted: for |
print.DMR for printing, plot.DMR for plotting, coef.DMR for extracting coefficients and predict.DMR for prediction.
## DMR for linear regression data(miete) ytr <- miete[1:1500,1] Xtr <- miete[1:1500,-1] Xte <- miete[1501:2053,-1] m1 <- DMR(Xtr, ytr) print(m1) plot(m1) g <- gic.DMR(m1, c = 2.5) plot(g) coef(m1, df = g$df.min) ypr <- predict(m1, newx = Xte, df = g$df.min) ## DMR for logistic regression # notice that only part of dataset promoter was used: DMR works only if p<n, for p>=n use DMRnet data(promoter) ytr <- factor(promoter[1:80,1]) Xtr <- promoter[1:80,2:11] Xte <- promoter[81:106,2:11] m2 <- DMR(Xtr, ytr, family = "binomial") print(m2) plot(m2) g <- gic.DMR(m2, c = 2) plot(g) coef(m2, df = g$df.min) ypr <- predict(m2, newx = Xte, df = g$df.min)## DMR for linear regression data(miete) ytr <- miete[1:1500,1] Xtr <- miete[1:1500,-1] Xte <- miete[1501:2053,-1] m1 <- DMR(Xtr, ytr) print(m1) plot(m1) g <- gic.DMR(m1, c = 2.5) plot(g) coef(m1, df = g$df.min) ypr <- predict(m1, newx = Xte, df = g$df.min) ## DMR for logistic regression # notice that only part of dataset promoter was used: DMR works only if p<n, for p>=n use DMRnet data(promoter) ytr <- factor(promoter[1:80,1]) Xtr <- promoter[1:80,2:11] Xte <- promoter[81:106,2:11] m2 <- DMR(Xtr, ytr, family = "binomial") print(m2) plot(m2) g <- gic.DMR(m2, c = 2) plot(g) coef(m2, df = g$df.min) ypr <- predict(m2, newx = Xte, df = g$df.min)
Fits a path of linear (family="gaussian") or logistic (family="binomial") regression models, where models are subsets of continuous predictors and partitions of levels of factors in X. Works even if p>=n (the number of observations is greater than the number of columns in the model matrix).
DMRnet( X, y, family = "gaussian", o = 5, nlambda = 100, lam = 10^(-7), interc = TRUE, maxp = ifelse(family == "gaussian", ceiling(length(y)/2), ceiling(length(y)/4)), lambda = NULL, algorithm = "DMRnet", clust.method = ifelse(algorithm == "glamer", "single", "complete") )DMRnet( X, y, family = "gaussian", o = 5, nlambda = 100, lam = 10^(-7), interc = TRUE, maxp = ifelse(family == "gaussian", ceiling(length(y)/2), ceiling(length(y)/4)), lambda = NULL, algorithm = "DMRnet", clust.method = ifelse(algorithm == "glamer", "single", "complete") )
X |
Input data frame; each row is an observation vector; each column can be numerical or integer for a continuous predictor or a factor for a categorical predictor. |
y |
Response variable; Numerical for |
family |
Response type; one of: |
o |
Parameter of the group lasso screening step, described in Details, the default value is 5. |
nlambda |
Parameter of the group lasso screening step, described in Details, the default value is 100. |
lam |
The amount of penalization in ridge regression (used for logistic regression in order to allow for parameter estimation in linearly separable setups) or the amount of matrix regularization in case of linear regression. Used only for numerical reasons. The default is 1e-7. |
interc |
Should intercept(s) be fitted (the default, |
maxp |
Maximal number of parameters of the model, smaller values result in quicker computation |
lambda |
Explicitly provided net of lambda values for the group lasso screening step, described in Details. If provided, it overrides the value of |
algorithm |
The algorithm to be used; for partition selection (merging levels) use one of: |
clust.method |
Clustering method used for partitioning levels of factors; see function hclust in package stats for details. |
DMRnet algorithm is a generalization of DMR to high-dimensional data.
It uses a screening step in order to decrease the problem to p<n and then uses DMR subsequently.
The screening is done with the group lasso algorithm implemented in the grpreg package.
First, the group lasso for the problem is solved for nlambda values of lambda parameter, or for the net of lambda values (if lambda is explicitly provided).
Next, for each value of lambda, the scaled nonzero second norms of the groups' coefficients are sorted in decreasing order.
Finally, the first i over o fraction of the groups with the largest nonzero values are chosen for further analysis, i = 1,2,...,o-1.
E.g., if o=5, first 1/5, first 2/5,..., 4/5 groups with the largest scaled nonzero second norm of coefficients are chosen.
The final path of models is chosen by minimizing the likelihood of the models for the number of parameters df equal to 1,...,l<=maxp for some integer l. Note that, in contrast to DMR, the models on the path need not to be nested.
An object with S3 class "DMR", which is a list with the ingredients:
beta |
Matrix p times l of estimated parameters; each column corresponds to a model on the nested path having from l to 1 parameter (denoted as df). |
df |
Vector of degrees of freedom; from l to 1. |
rss/loglik |
Measure of fit for the nested models: rss (residual sum of squares) is returned for |
n |
Number of observations. |
levels.listed |
Minimal set of levels of respective factors present in data. |
lambda |
The net of lambda values used in the screening step. |
arguments |
List of the chosen arguments from the function call. |
interc |
If the intercept was fitted: value of parameter |
print.DMR for printing, plot.DMR for plotting, coef.DMR for extracting coefficients and predict.DMR for prediction.
## DMRnet for linear regression data(miete) ytr <- miete[1:200,1] Xtr <- miete[1:200,-1] Xte <- miete[201:250,-1] m1 <- DMRnet(Xtr, ytr) print(m1) plot(m1) g <- gic.DMR(m1, c = 2.5) plot(g) coef(m1, df = g$df.min) ypr <- predict(m1, newx = Xte, df = g$df.min) ## DMRnet for logistic regression data(promoter) ytr <- factor(promoter[1:70,1]) Xtr <- promoter[1:70,-1] Xte <- promoter[71:106,-1] m2 <- DMRnet(Xtr, ytr, family = "binomial") print(m2) plot(m2) g <- gic.DMR(m2, c = 2) plot(g) coef(m2, df = g$df.min) ypr <- predict(m2, newx = Xte, df = g$df.min) ## PDMR for linear regression data(miete) ytr <- miete[1:200,1] Xtr <- miete[1:200,-1] Xte <- miete[201:250,-1] m1 <- DMRnet(Xtr, ytr, algorithm="PDMR") print(m1) plot(m1) g <- gic.DMR(m1, c = 2.5) plot(g) coef(m1, df = g$df.min) ypr <- predict(m1, newx = Xte, df = g$df.min)## DMRnet for linear regression data(miete) ytr <- miete[1:200,1] Xtr <- miete[1:200,-1] Xte <- miete[201:250,-1] m1 <- DMRnet(Xtr, ytr) print(m1) plot(m1) g <- gic.DMR(m1, c = 2.5) plot(g) coef(m1, df = g$df.min) ypr <- predict(m1, newx = Xte, df = g$df.min) ## DMRnet for logistic regression data(promoter) ytr <- factor(promoter[1:70,1]) Xtr <- promoter[1:70,-1] Xte <- promoter[71:106,-1] m2 <- DMRnet(Xtr, ytr, family = "binomial") print(m2) plot(m2) g <- gic.DMR(m2, c = 2) plot(g) coef(m2, df = g$df.min) ypr <- predict(m2, newx = Xte, df = g$df.min) ## PDMR for linear regression data(miete) ytr <- miete[1:200,1] Xtr <- miete[1:200,-1] Xte <- miete[201:250,-1] m1 <- DMRnet(Xtr, ytr, algorithm="PDMR") print(m1) plot(m1) g <- gic.DMR(m1, c = 2.5) plot(g) coef(m1, df = g$df.min) ypr <- predict(m1, newx = Xte, df = g$df.min)
Computes values of Generalized Information Criterion for the entire sequence of models from a DMR object.
gic.DMR( x, c = ifelse(x$arguments$family == "gaussian", constants()$RIC_gaussian_constant, constants()$RIC_binomial_constant) )gic.DMR( x, c = ifelse(x$arguments$family == "gaussian", constants()$RIC_gaussian_constant, constants()$RIC_binomial_constant) )
x |
Fitted |
c |
Parameter controlling amount of penalization for complexity of the model in the generalized information criterion (GIC). For linear regression GIC for model M is defined as
where
where |
An object of class "gic.DMR" is returned, which is a list with the ingredients of the gic fit.
df (number of parameters) for the model with minimal GIC.
Fitted DMR object.
Vector of GIC values for the entire sequence of models.
plot.gic.DMR for plotting, coef.gic.DMR for extracting coefficients and predict.gic.DMR for prediction.
data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) (g <- gic.DMR(m, c = 2.5))data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) (g <- gic.DMR(m, c = 2.5))
The miete data contains the rent index for Munich in 2003.
data(miete)data(miete)
A data frame with 2053 observations on the following 12 variables.
Rent in euros.
Special furniture in bathroom, yes = 1, no = 0.
Bathroom with tiles, yes = 0, no = 1.
Municipality.
Upmarket kitchen, yes = 1, no = 0.
Number of rooms.
Best address, yes = 1, no = 0.
Good address, yes = 1, no =0.
Warm water, yes = 0, no = 1.
Central heating, yes = 0, no = 1.
Year of construction.
Living space in square meter.
Fahrmeir, L., Künstler, R., Pigeot, I., Tutz, G. (2004) Statistik: der Weg zur Datenanalyse. 5. Auflage, Berlin: Springer-Verlag.
data(miete) summary(miete)data(miete) summary(miete)
Plots cross-validated error values from a cv.DMR object.
## S3 method for class 'cv.DMR' plot(x, ...)## S3 method for class 'cv.DMR' plot(x, ...)
x |
Fitted |
... |
Further arguments passed to or from other methods. |
Produces a plot of cross-validated error values for the entire sequence of models from the fitted cv.DMR object. The horizontal level indicating separation of one standard deviation from the minimum error is indicated with a blue dashed line. The df.min (the smallest model minimizing the cross-validated error) and df.1se (the smallest model falling under the blue dashed line) are marked with red and blue points, respectively.
## cv.DMR for linear regression set.seed(13) data(miete) y <- miete$rent X <- miete$area cv = cv.DMR(X,y) plot(cv)## cv.DMR for linear regression set.seed(13) data(miete) y <- miete$rent X <- miete$area cv = cv.DMR(X,y) plot(cv)
Plots coefficients from a DMR object.
## S3 method for class 'DMR' plot(x, ...)## S3 method for class 'DMR' plot(x, ...)
x |
Fitted |
... |
Further arguments passed to or from other methods. |
Produces a coefficient profile plot of the coefficient paths for a fitted DMR object.
data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) plot(m)data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) plot(m)
Plots gic values from a gic.DMR object.
## S3 method for class 'gic.DMR' plot(x, ...)## S3 method for class 'gic.DMR' plot(x, ...)
x |
Fitted |
... |
Further arguments passed to or from other methods. |
Produces a plot of Generalized Information Criterion for the entire sequence of models from the fitted gic.DMR object.
data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) g <- gic.DMR(m, c = 2.5) plot(g)data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) g <- gic.DMR(m, c = 2.5) plot(g)
Makes predictions from a cv.DMR object (for the model with minimal cross-validated error /the default/ or the smallest model falling under the upper curve of a prediction error plus one standard deviation).
## S3 method for class 'cv.DMR' predict( object, newx, type = "link", md = "df.min", unknown.factor.levels = "error", ... )## S3 method for class 'cv.DMR' predict( object, newx, type = "link", md = "df.min", unknown.factor.levels = "error", ... )
object |
Fitted cv.DMR object. |
newx |
Data frame of new values for |
type |
One of: |
md |
Value of the model dimension parameter at which predictions are required. The default is |
unknown.factor.levels |
The way of handling factor levels in test data not seen while training a model. One of |
... |
Further arguments passed to or from other methods. |
Similar to other predict methods, this function predicts fitted values from a fitted cv.DMR object.
Vector of predictions.
## cv.DMR for linear regression set.seed(13) data(miete) ytr <- miete$rent[1:1500] Xtr <- miete$area[1:1500] Xte <- miete$area[1501:2053] cv <- cv.DMR(Xtr, ytr) print(cv) plot(cv) coef(cv) ypr <- predict(cv, newx = Xte)## cv.DMR for linear regression set.seed(13) data(miete) ytr <- miete$rent[1:1500] Xtr <- miete$area[1:1500] Xte <- miete$area[1501:2053] cv <- cv.DMR(Xtr, ytr) print(cv) plot(cv) coef(cv) ypr <- predict(cv, newx = Xte)
Makes predictions from a DMR object.
## S3 method for class 'DMR' predict( object, newx, df = NULL, type = "link", unknown.factor.levels = "error", ... )## S3 method for class 'DMR' predict( object, newx, df = NULL, type = "link", unknown.factor.levels = "error", ... )
object |
Fitted |
newx |
Data frame of new values for |
df |
Number of parameters in the model for which predictions are required. Default is the entire sequence of models for df=1 to df=p. |
type |
One of: |
unknown.factor.levels |
The way of handling factor levels in test data not seen while training a model. One of |
... |
Further arguments passed to or from other methods. |
Similar to other predict methods, this function predicts fitted values from a fitted DMR object.
Vector or matrix of predictions.
data(miete) ytr <- miete[1:1500,1] Xtr <- miete[1:1500,-1] Xte <- miete[1501:2053,-1] m <- DMR(Xtr, ytr) ypr <- predict(m, newx = Xte, df = 11)data(miete) ytr <- miete[1:1500,1] Xtr <- miete[1:1500,-1] Xte <- miete[1501:2053,-1] m <- DMR(Xtr, ytr) ypr <- predict(m, newx = Xte, df = 11)
Makes predictions from a gic.DMR object (for the model with minimal GIC).
## S3 method for class 'gic.DMR' predict(object, newx, type = "link", unknown.factor.levels = "error", ...)## S3 method for class 'gic.DMR' predict(object, newx, type = "link", unknown.factor.levels = "error", ...)
object |
Fitted |
newx |
Data frame of new values for |
type |
One of: |
unknown.factor.levels |
The way of handling factor levels in test data not seen while training a model. One of |
... |
Further arguments passed to or from other methods. |
Similar to other predict methods, this function predicts fitted values from a fitted gic.DMR object for the model with minimal GIC.
Vector of predictions.
data(miete) ytr <- miete[1:1500,1] Xtr <- miete[1:1500,-1] Xte <- miete[1501:2053,-1] m <- DMR(Xtr, ytr) g <- gic.DMR(m, c = 2.5) ypr <- predict(g, newx = Xte)data(miete) ytr <- miete[1:1500,1] Xtr <- miete[1:1500,-1] Xte <- miete[1501:2053,-1] m <- DMR(Xtr, ytr) g <- gic.DMR(m, c = 2.5) ypr <- predict(g, newx = Xte)
Prints a DMR object.
## S3 method for class 'DMR' print(x, ...)## S3 method for class 'DMR' print(x, ...)
x |
Fitted |
... |
Further arguments passed to or from other methods. |
Print a summary of the DMR path at each step along the path.
The summary is silently returned.
data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) print(m)data(miete) y <- miete[,1] X <- miete[,-1] m <- DMR(X, y) print(m)
It consists of E. coli promoter gene sequences starting at position -50 (p-50) and ending at position +7 (p7). Each of these 57 Fields is filled by one of a, g, t, c. The task is to recognize promoters, which are genetic regions which initiate the first step in the expression of adjacent genes (transcription). There are 53 promoters and 53 non-promoter sequences.
data(promoter)data(promoter)
A data frame with 106 observations on the following 58 variables.
One of 1/0, indicating the class (1 = promoter).
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
Sequence; filled by one of a, g, t, c.
UCI machine learning repository: promoter
Towell, G., Shavlik, J., Noordewier, M. Refinement of approximate domain theories by knowledge-based neural networks. In Proceedings of the eighth National conference on Artificial intelligence, pages 861-866. Boston, MA, 1990.
data(promoter) summary(promoter)data(promoter) summary(promoter)