Package 'DALEX'

Description

Datasets apartments and apartments_test are artificial, generated form the same model. Structure of the dataset is copied from real dataset from PBImisc package, but they were generated in a way to mimic effect of Anscombe quartet for complex black box models.

Usage

data(apartments)

Format

a data frame with 1000 rows and 6 columns

Details

• m2.price - price per square meter

• surface - apartment area in square meters

• n.rooms - number of rooms (correlated with surface)

• district - district in which apartment is located, factor with 10 levels

• floor - floor

• construction.date - construction year

Description

DrWhy color palettes for ggplot objects

Usage

colors_discrete_drwhy(n = 2)

colors_diverging_drwhy()

colors_breakdown_drwhy()

Arguments

Value

color palette as vector of charactes

Description

Two datasets of characteristics of patients infected with COVID. It is important to note that these are not real patient data. This is simulated data, generated to have relationships consistent with real data (obtained from NIH), but the data itself is not real. Fortunately, they are sufficient for the purposes of our exercise.

Usage

data(covid_summer)
data(covid_spring)

Format

a data frame with 10 000 rows each and 12 columns

Details

The data is divided into two sets covid_spring and covid_summer. The first is acquired in spring 2020 and will be used as training data while the second dataset is acquired in summer and will be used for validation. In machine learning, model validation is performed on a separate data set. This controls the risk of overfitting an elastic model to the data. If we do not have a separate set then it is generated using cross-validation, out of sample or out of time techniques.

It contains 20 000 rows related fo COVID mortality. it contains 11 variables such as: Gender, Age, Cardiovascular.Diseases, Diabetes, Neurological.Diseases, Kidney.Diseases.

Source: https://github.com/BetaAndBit/RML

Source

https://github.com/BetaAndBit/RML

Description

Datasets dragons and dragons_test are artificial, generated form the same ground truth model, but with sometimes different data distribution.

Usage

data(dragons)

Format

a data frame with 2000 rows and 8 columns

Details

Values are generated in a way to: - have nonlinearity in year_of_birth and height - have concept drift in the test set

• year_of_birth - year in which the dragon was born. Negative year means year BC, eg: -1200 = 1201 BC

• year_of_discovery - year in which the dragon was found.

• height - height of the dragon in yards.

• weight - weight of the dragon in tons.

• scars - number of scars.

• colour - colour of the dragon.

• number_of_lost_teeth - number of teeth that the dragon lost.

• life_length - life length of the dragon.

Description

Black-box models may have very different structures. This function creates a unified representation of a model, which can be further processed by functions for explanations.

Usage

explain.default(
  model,
  data = NULL,
  y = NULL,
  predict_function = NULL,
  predict_function_target_column = NULL,
  residual_function = NULL,
  weights = NULL,
  ...,
  label = NULL,
  verbose = TRUE,
  precalculate = TRUE,
  colorize = !isTRUE(getOption("knitr.in.progress")),
  model_info = NULL,
  type = NULL
)

explain(
  model,
  data = NULL,
  y = NULL,
  predict_function = NULL,
  predict_function_target_column = NULL,
  residual_function = NULL,
  weights = NULL,
  ...,
  label = NULL,
  verbose = TRUE,
  precalculate = TRUE,
  colorize = !isTRUE(getOption("knitr.in.progress")),
  model_info = NULL,
  type = NULL
)

Arguments

Details

Please NOTE that the model is the only required argument. But some explanations may expect that other arguments will be provided too.

Value

An object of the class explainer.

It's a list with the following fields:

• model the explained model.

• data the dataset used for training.

• y response for observations from data.

• weights sample weights for data. NULL if weights are not specified.

• y_hat calculated predictions.

• residuals calculated residuals.

• predict_function function that may be used for model predictions, shall return a single numerical value for each observation.

• residual_function function that returns residuals, shall return a single numerical value for each observation.

• class class/classes of a model.

• label label of explainer.

• model_info named list contating basic information about model, like package, version of package and type.

References

Explanatory Model Analysis. Explore, Explain and Examine Predictive Models. https://ema.drwhy.ai/

Examples

# simple explainer for regression problem
aps_lm_model4 <- lm(m2.price ~., data = apartments)
aps_lm_explainer4 <- explain(aps_lm_model4, data = apartments, label = "model_4v")
aps_lm_explainer4

# various parameters for the explain function
# all defaults
aps_lm <- explain(aps_lm_model4)

# silent execution
aps_lm <- explain(aps_lm_model4, verbose = FALSE)

# set target variable
aps_lm <- explain(aps_lm_model4, data = apartments, label = "model_4v", y = apartments$m2.price)
aps_lm <- explain(aps_lm_model4, data = apartments, label = "model_4v", y = apartments$m2.price,
                                   predict_function = predict)


# user provided predict_function
aps_ranger <- ranger::ranger(m2.price~., data = apartments, num.trees = 50)
custom_predict <- function(X.model, newdata) {
   predict(X.model, newdata)$predictions
}
aps_ranger_exp <- explain(aps_ranger, data = apartments, y = apartments$m2.price,
                          predict_function = custom_predict)


# user provided residual_function
aps_ranger <- ranger::ranger(m2.price~., data = apartments, num.trees = 50)
custom_residual <- function(X.model, newdata, y, predict_function) {
   abs(y - predict_function(X.model, newdata))
}
aps_ranger_exp <- explain(aps_ranger, data = apartments,
                          y = apartments$m2.price,
                          residual_function = custom_residual)

# binary classification
titanic_ranger <- ranger::ranger(as.factor(survived)~., data = titanic_imputed, num.trees = 50,
                                 probability = TRUE)
# keep in mind that for binary classification y parameter has to be numeric  with 0 and 1 values
titanic_ranger_exp <- explain(titanic_ranger, data = titanic_imputed, y = titanic_imputed$survived)

# multiclass task
hr_ranger <- ranger::ranger(status~., data = HR, num.trees = 50, probability = TRUE)
# keep in mind that for multiclass y parameter has to be a factor,
# with same levels as in training data
hr_ranger_exp <- explain(hr_ranger, data = HR, y = HR$status)

# set model_info
model_info <- list(package = "stats", ver = "3.6.2", type = "regression")
aps_lm_model4 <- lm(m2.price ~., data = apartments)
aps_lm_explainer4 <- explain(aps_lm_model4, data = apartments, label = "model_4v",
                             model_info = model_info)

# simple function
aps_fun <- function(x) 58*x$surface
aps_fun_explainer <- explain(aps_fun, data = apartments, y = apartments$m2.price, label="sfun")
model_performance(aps_fun_explainer)

# set model_info
model_info <- list(package = "stats", ver = "3.6.2", type = "regression")
aps_lm_model4 <- lm(m2.price ~., data = apartments)
aps_lm_explainer4 <- explain(aps_lm_model4, data = apartments, label = "model_4v",
                             model_info = model_info)

aps_lm_explainer4 <- explain(aps_lm_model4, data = apartments, label = "model_4v",
                             weights = as.numeric(apartments$construction.year > 2000))

# more complex model
library("ranger")
aps_ranger_model4 <- ranger(m2.price ~., data = apartments, num.trees = 50)
aps_ranger_explainer4 <- explain(aps_ranger_model4, data = apartments, label = "model_ranger")
aps_ranger_explainer4

Description

The fifa dataset is a preprocessed players_20.csv dataset which comes as a part of "FIFA 20 complete player dataset" at Kaggle.

Usage

data(fifa)

Format

a data frame with 5000 rows, 42 columns and rownames

Details

It contains 5000 'overall' best players and 43 variables. These are:

• short_name (rownames)

• nationality of the player (not used in modeling)

• overall, potential, value_eur, wage_eur (4 potential target variables)

• age, height, weight, attacking skills, defending skills, goalkeeping skills (37 variables)

It is advised to leave only one target variable for modeling.

Source: https://www.kaggle.com/stefanoleone992/fifa-20-complete-player-dataset

All transformations:

1. take 43 columns: [3, 5, 7:9, 11:14, 45:78] (R indexing)

2. take rows with value_eur > 0

3. convert short_name to ASCII

4. remove rows with duplicated short_name (keep first)

5. sort rows on overall and take top 5000

6. set short_name column as rownames

7. transform nationality to factor

8. reorder columns

Source

The players_20.csv dataset was downloaded from the Kaggle site and went through few transformations. The complete dataset was obtained from https://www.kaggle.com/stefanoleone992/fifa-20-complete-player-dataset#players_20.csv on January 1, 2020.

Description

The yardstick package provides many auxiliary functions for calculating the predictive performance of the model. However, they have an interface that is consistent with the tidyverse philosophy. The loss_yardstick function adapts loss functions from the yardstick package to functions understood by DALEX. Type compatibility for y-values and for predictions must be guaranteed by the user.

Usage

get_loss_yardstick(loss, reverse = FALSE, reference = 1)

loss_yardstick(loss, reverse = FALSE, reference = 1)

Arguments

Value

loss function that can be used in the model_parts function

Examples

titanic_glm_model <- glm(survived~., data = titanic_imputed, family = "binomial")
 explainer_glm <- DALEX::explain(titanic_glm_model,
                                 data = titanic_imputed[,-8],
                                 y = factor(titanic_imputed$survived))
 # See the 'How to use DALEX with the yardstick package' vignette
 # which explains this model with measures implemented in the 'yardstick' package

Description

The happiness_train and happiness_test datasets are generated based on the "World Happiness Report" at Kaggle https://www.kaggle.com/datasets/unsdsn/world-happiness.

Usage

data(happiness_train)
data(happiness_test)

Format

two data frames with total 781 rows, 7 columns ech and rownames

Details

It contains data for 781 countries and 7 variables. These are:

• score - Happiness score

• gdp_per_capita - GDP per capita

• social_support - Social support

• healthy_life_expectancy - Healthy life expectancy

• freedom_life_choices - Freedom to make life choices

• generosity - Generosity

• perceptions_of_corruption - Perceptions of corruption

Source

World Happiness Report data https://worldhappiness.report/

Description

Datasets HR and HR_test are artificial, generated form the same model. Structure of the dataset is based on a real data, from Human Resources department with information which employees were promoted, which were fired.

Usage

data(HR)

Format

a data frame with 10000 rows and 6 columns

Details

Values are generated in a way to: - have interaction between age and gender for the 'fired' variable - have non monotonic relation for the salary variable - have linear effects for hours and evaluation.

• gender - gender of an employee.

• age - age of an employee in the moment of evaluation.

• hours - average number of working hours per week.

• evaluation - evaluation in the scale 2 (bad) - 5 (very good).

• salary - level of salary in the scale 0 (lowest) - 5 (highest).

• status - target variable, either 'fired' or 'promoted' or 'ok'.

Description

By default 'heavy' dependencies are not installed along DALEX. This function silently install all required packages.

Usage

install_dependencies(packages = c("ingredients", "iBreakDown", "ggpubr"))

Arguments

Description

Calculate Loss Functions

Usage

loss_cross_entropy(observed, predicted, p_min = 1e-04, na.rm = TRUE)

loss_sum_of_squares(observed, predicted, na.rm = TRUE)

loss_root_mean_square(observed, predicted, na.rm = TRUE)

loss_accuracy(observed, predicted, na.rm = TRUE)

loss_one_minus_accuracy(observed, predicted, cutoff = 0.5, na.rm = TRUE)

get_loss_one_minus_accuracy(cutoff = 0.5, na.rm = TRUE)

loss_one_minus_auc(observed, predicted)

get_loss_default(x)

loss_default(x)

Arguments

Value

numeric - value of the loss function

Examples

library("ranger")
titanic_ranger_model <- ranger(survived~., data = titanic_imputed, num.trees = 50,
                               probability = TRUE)
loss_one_minus_auc(titanic_imputed$survived, yhat(titanic_ranger_model, titanic_imputed))

HR_ranger_model_multi <- ranger(status~., data = HR, num.trees = 50, probability = TRUE)
loss_cross_entropy(as.numeric(HR$status), yhat(HR_ranger_model_multi, HR))

Description

This function performs model diagnostic of residuals. Residuals are calculated and plotted against predictions, true y values or selected variables. Find information how to use this function here: https://ema.drwhy.ai/residualDiagnostic.html.

Usage

model_diagnostics(explainer, variables = NULL, ...)

Arguments

Value

An object of the class model_diagnostics. It's a data frame with residuals and selected variables.

References

Explanatory Model Analysis. Explore, Explain and Examine Predictive Models. https://ema.drwhy.ai/

Examples

library(DALEX)
apartments_lm_model <- lm(m2.price ~ ., data = apartments)
explainer_lm <- explain(apartments_lm_model,
                         data = apartments,
                         y = apartments$m2.price)
diag_lm <- model_diagnostics(explainer_lm)
diag_lm
plot(diag_lm)

library("ranger")
apartments_ranger_model <- ranger(m2.price ~ ., data = apartments)
explainer_ranger <- explain(apartments_ranger_model,
                         data = apartments,
                         y = apartments$m2.price)
diag_ranger <- model_diagnostics(explainer_ranger)
diag_ranger
plot(diag_ranger)
plot(diag_ranger, diag_lm)
plot(diag_ranger, diag_lm, variable = "y")
plot(diag_ranger, diag_lm, variable = "construction.year")
plot(diag_ranger, variable = "y", yvariable = "y_hat")
plot(diag_ranger, variable = "y", yvariable = "abs_residuals")
plot(diag_ranger, variable = "ids")

Description

This generic function let user extract base information about model. The function returns a named list of class model_info that contain about package of model, version and task type. For wrappers like mlr or caret both, package and wrapper inforamtion are stored

Usage

model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'lm'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'randomForest'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'svm'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'glm'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'lrm'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'glmnet'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'cv.glmnet'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'ranger'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'gbm'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'model_fit'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'train'
model_info(model, is_multiclass = FALSE, ...)

## S3 method for class 'rpart'
model_info(model, is_multiclass = FALSE, ...)

## Default S3 method:
model_info(model, is_multiclass = FALSE, ...)

Arguments

Details

Currently supported packages are:

• class cv.glmnet and glmnet - models created with glmnet package

• class glm - generalized linear models

• class lrm - models created with rms package,

• class model_fit - models created with parsnip package

• class lm - linear models created with stats::lm

• class ranger - models created with ranger package

• class randomForest - random forest models created with randomForest package

• class svm - support vector machines models created with the e1071 package

• class train - models created with caret package

• class gbm - models created with gbm package

Value

A named list of class model_info

Examples

aps_lm_model4 <- lm(m2.price ~., data = apartments)
model_info(aps_lm_model4)


library("ranger")
model_regr_rf <- ranger::ranger(status~., data = HR, num.trees = 50, probability = TRUE)
model_info(model_regr_rf, is_multiclass = TRUE)

Description

From DALEX version 1.0 this function calls the feature_importance Find information how to use this function here: https://ema.drwhy.ai/featureImportance.html.

Usage

model_parts(
  explainer,
  loss_function = get_loss_default(explainer$model_info$type),
  ...,
  type = "variable_importance",
  N = n_sample,
  n_sample = 1000
)

Arguments

Value

An object of the class feature_importance. It's a data frame with calculated average response.

References

Explanatory Model Analysis. Explore, Explain and Examine Predictive Models. https://ema.drwhy.ai/

Examples

# regression

library("ranger")
apartments_ranger_model <- ranger(m2.price~., data = apartments, num.trees = 50)
explainer_ranger  <- explain(apartments_ranger_model, data = apartments[,-1],
                             y = apartments$m2.price, label = "Ranger Apartments")
model_parts_ranger_aps <- model_parts(explainer_ranger, type = "raw")
head(model_parts_ranger_aps, 8)
plot(model_parts_ranger_aps)

# binary classification

titanic_glm_model <- glm(survived~., data = titanic_imputed, family = "binomial")
explainer_glm_titanic <- explain(titanic_glm_model, data = titanic_imputed[,-8],
                         y = titanic_imputed$survived)
logit <- function(x) exp(x)/(1+exp(x))
custom_loss <- function(observed, predicted){
   sum((observed - logit(predicted))^2)
}
attr(custom_loss, "loss_name") <- "Logit residuals"
model_parts_glm_titanic <- model_parts(explainer_glm_titanic, type = "raw",
                                       loss_function = custom_loss)
head(model_parts_glm_titanic, 8)
plot(model_parts_glm_titanic)

# multilabel classification

HR_ranger_model_HR <- ranger(status~., data = HR, num.trees = 50,
                               probability = TRUE)
explainer_ranger_HR  <- explain(HR_ranger_model_HR, data = HR[,-6],
                             y = HR$status, label = "Ranger HR")
model_parts_ranger_HR <- model_parts(explainer_ranger_HR, type = "raw")
head(model_parts_ranger_HR, 8)
plot(model_parts_ranger_HR)

Description

Function model_performance() calculates various performance measures for classification and regression models. For classification models following measures are calculated: F1, accuracy, recall, precision and AUC. For regression models following measures are calculated: mean squared error, R squared, median absolute deviation.

Usage

model_performance(explainer, ..., cutoff = 0.5)

Arguments

Value

An object of the class model_performance.

It's a list with following fields:

• residuals - data frame that contains residuals for each observation

• measures - list with calculated measures that are dedicated for the task, whether it is regression, binary classification or multiclass classification.

• type - character that specifies type of the task.

References

Explanatory Model Analysis. Explore, Explain, and Examine Predictive Models. https://ema.drwhy.ai/

Examples

# regression

library("ranger")
apartments_ranger_model <- ranger(m2.price~., data = apartments, num.trees = 50)
explainer_ranger_apartments  <- explain(apartments_ranger_model, data = apartments[,-1],
                             y = apartments$m2.price, label = "Ranger Apartments")
model_performance_ranger_aps <- model_performance(explainer_ranger_apartments )
model_performance_ranger_aps
plot(model_performance_ranger_aps)
plot(model_performance_ranger_aps, geom = "boxplot")
plot(model_performance_ranger_aps, geom = "histogram")

# binary classification

titanic_glm_model <- glm(survived~., data = titanic_imputed, family = "binomial")
explainer_glm_titanic <- explain(titanic_glm_model, data = titanic_imputed[,-8],
                         y = titanic_imputed$survived)
model_performance_glm_titanic <- model_performance(explainer_glm_titanic)
model_performance_glm_titanic
plot(model_performance_glm_titanic)
plot(model_performance_glm_titanic, geom = "boxplot")
plot(model_performance_glm_titanic, geom = "histogram")

# multilabel classification

HR_ranger_model <- ranger(status~., data = HR, num.trees = 50,
                               probability = TRUE)
explainer_ranger_HR  <- explain(HR_ranger_model, data = HR[,-6],
                             y = HR$status, label = "Ranger HR")
model_performance_ranger_HR <- model_performance(explainer_ranger_HR)
model_performance_ranger_HR
plot(model_performance_ranger_HR)
plot(model_performance_ranger_HR, geom = "boxplot")
plot(model_performance_ranger_HR, geom = "histogram")

Description

This function calculates explanations on a dataset level set that explore model response as a function of selected variables. The explanations can be calulated as Partial Dependence Profile or Accumulated Local Dependence Profile. Find information how to use this function here: https://ema.drwhy.ai/partialDependenceProfiles.html. The variable_profile function is a copy of model_profile.

Usage

model_profile(
  explainer,
  variables = NULL,
  N = 100,
  ...,
  groups = NULL,
  k = NULL,
  center = TRUE,
  type = "partial"
)

variable_profile(
  explainer,
  variables = NULL,
  N = 100,
  ...,
  groups = NULL,
  k = NULL,
  center = TRUE,
  type = "partial"
)

single_variable(explainer, variable, type = "pdp", ...)

Arguments

Details

Underneath this function calls the partial_dependence or accumulated_dependence functions from the ingredients package.

Value

An object of the class model_profile. It's a data frame with calculated average model responses.

References

Explanatory Model Analysis. Explore, Explain, and Examine Predictive Models. https://ema.drwhy.ai/

Examples

titanic_glm_model <- glm(survived~., data = titanic_imputed, family = "binomial")
explainer_glm <- explain(titanic_glm_model, data = titanic_imputed)
model_profile_glm_fare <- model_profile(explainer_glm, "fare")
plot(model_profile_glm_fare)

 
library("ranger")
titanic_ranger_model <- ranger(survived~., data = titanic_imputed, num.trees = 50,
                               probability = TRUE)
explainer_ranger  <- explain(titanic_ranger_model, data = titanic_imputed)
model_profile_ranger <- model_profile(explainer_ranger)
plot(model_profile_ranger, geom = "profiles")

model_profile_ranger_1 <- model_profile(explainer_ranger, type = "partial",
                                        variables = c("age", "fare"))
plot(model_profile_ranger_1 , variables = c("age", "fare"), geom = "points")

model_profile_ranger_2  <- model_profile(explainer_ranger, type = "partial", k = 3)
plot(model_profile_ranger_2 , geom = "profiles")

model_profile_ranger_3  <- model_profile(explainer_ranger, type = "partial", groups = "gender")
plot(model_profile_ranger_3 , geom = "profiles")

model_profile_ranger_4  <- model_profile(explainer_ranger, type = "accumulated")
plot(model_profile_ranger_4 , geom = "profiles")

# Multiple profiles
model_profile_ranger_fare <- model_profile(explainer_ranger, "fare")
plot(model_profile_ranger_fare, model_profile_glm_fare)

Description

Plot List of Explanations

Usage

## S3 method for class 'list'
plot(x, ...)

Arguments

Value

An object of the class ggplot.

Examples

library("ranger")
titanic_ranger_model <- ranger(survived~., data = titanic_imputed, num.trees = 50,
                               probability = TRUE)
explainer_ranger  <- explain(titanic_ranger_model, data = titanic_imputed[,-8],
                             y = titanic_imputed$survived)
mp_ranger <- model_performance(explainer_ranger)

titanic_ranger_model2 <- ranger(survived~gender + fare, data = titanic_imputed,
                                num.trees = 50, probability = TRUE)
explainer_ranger2  <- explain(titanic_ranger_model2, data = titanic_imputed[,-8],
                              y = titanic_imputed$survived,
                              label = "ranger2")
mp_ranger2 <- model_performance(explainer_ranger2)

plot(list(mp_ranger, mp_ranger2), geom = "prc")
plot(list(mp_ranger, mp_ranger2), geom = "roc")
tmp <- list(mp_ranger, mp_ranger2)
names(tmp) <- c("ranger", "ranger2")
plot(tmp)

Description

Plot Dataset Level Model Diagnostics

Usage

## S3 method for class 'model_diagnostics'
plot(x, ..., variable = "y_hat", yvariable = "residuals", smooth = TRUE)

Arguments

Value

an object of the class model_diagnostics_explainer.

Examples

apartments_lm_model <- lm(m2.price ~ ., data = apartments)
explainer_lm <- explain(apartments_lm_model,
                         data = apartments,
                         y = apartments$m2.price)
diag_lm <- model_diagnostics(explainer_lm)
diag_lm
plot(diag_lm)

library("ranger")
apartments_ranger_model <- ranger(m2.price ~ ., data = apartments)
explainer_ranger <- explain(apartments_ranger_model,
                         data = apartments,
                         y = apartments$m2.price)
diag_ranger <- model_diagnostics(explainer_ranger)
diag_ranger
plot(diag_ranger)
plot(diag_ranger, diag_lm)
plot(diag_ranger, diag_lm, variable = "y")
plot(diag_ranger, diag_lm, variable = "construction.year")
plot(diag_ranger, variable = "y", yvariable = "y_hat")

Description

Plot Variable Importance Explanations

Usage

## S3 method for class 'model_parts'
plot(x, ...)

Arguments

Value

An object of the class ggplot.

Plot options

variable_importance

• max_vars maximal number of features to be included in the plot. default value is 10

• show_boxplots logical if TRUE (default) boxplot will be plotted to show permutation data.

• bar_width width of bars. By default 10

• desc_sorting logical. Should the bars be sorted descending? By default TRUE

• title the plot's title, by default 'Feature Importance'

• subtitle a character. Plot subtitle. By default NULL - then subtitle is set to "created for the XXX, YYY model", where XXX, YYY are labels of given explainers.

Description

Plot Dataset Level Model Performance Explanations

Usage

## S3 method for class 'model_performance'
plot(
  x,
  ...,
  geom = "ecdf",
  show_outliers = 0,
  ptlabel = "name",
  lossFunction = loss_function,
  loss_function = function(x) sqrt(mean(x^2))
)

Arguments

Value

An object of the class model_performance.

Examples

library("ranger")
titanic_ranger_model <- ranger(survived~., data = titanic_imputed, num.trees = 50,
                               probability = TRUE)
explainer_ranger  <- explain(titanic_ranger_model, data = titanic_imputed[,-8],
                             y = titanic_imputed$survived)
mp_ranger <- model_performance(explainer_ranger)
plot(mp_ranger)
plot(mp_ranger, geom = "boxplot", show_outliers = 1)

titanic_ranger_model2 <- ranger(survived~gender + fare, data = titanic_imputed,
                                num.trees = 50, probability = TRUE)
explainer_ranger2  <- explain(titanic_ranger_model2, data = titanic_imputed[,-8],
                              y = titanic_imputed$survived,
                              label = "ranger2")
mp_ranger2 <- model_performance(explainer_ranger2)
plot(mp_ranger, mp_ranger2, geom = "prc")
plot(mp_ranger, mp_ranger2, geom = "roc")
plot(mp_ranger, mp_ranger2, geom = "lift")
plot(mp_ranger, mp_ranger2, geom = "gain")
plot(mp_ranger, mp_ranger2, geom = "boxplot")
plot(mp_ranger, mp_ranger2, geom = "histogram")
plot(mp_ranger, mp_ranger2, geom = "ecdf")

titanic_glm_model <- glm(survived~., data = titanic_imputed, family = "binomial")
explainer_glm <- explain(titanic_glm_model, data = titanic_imputed[,-8],
                         y = titanic_imputed$survived, label = "glm",
                    predict_function = function(m,x) predict.glm(m,x,type = "response"))
mp_glm <- model_performance(explainer_glm)
plot(mp_glm)

titanic_lm_model <- lm(survived~., data = titanic_imputed)
explainer_lm <- explain(titanic_lm_model, data = titanic_imputed[,-8],
                        y = titanic_imputed$survived, label = "lm")
mp_lm <- model_performance(explainer_lm)
plot(mp_lm)

plot(mp_ranger, mp_glm, mp_lm)
plot(mp_ranger, mp_glm, mp_lm, geom = "boxplot")
plot(mp_ranger, mp_glm, mp_lm, geom = "boxplot", show_outliers = 1)

Description

Plot Dataset Level Model Profile Explanations

Usage

## S3 method for class 'model_profile'
plot(x, ..., geom = "aggregates")

Arguments

Value

An object of the class ggplot.

aggregates

• color a character. Either name of a color, or hex code for a color, or _label_ if models shall be colored, or _ids_ if instances shall be colored

• size a numeric. Size of lines to be plotted

• alpha a numeric between 0 and 1. Opacity of lines

• facet_ncol number of columns for the facet_wrap

• variables if not NULL then only variables will be presented

• title a character. Partial and accumulated dependence explainers have deafult value.

• subtitle a character. If NULL value will be dependent on model usage.

Examples

titanic_glm_model <- glm(survived~., data = titanic_imputed, family = "binomial")
explainer_glm <- explain(titanic_glm_model, data = titanic_imputed)
expl_glm <- model_profile(explainer_glm, "fare")
plot(expl_glm)

 
library("ranger")
titanic_ranger_model <- ranger(survived~., data = titanic_imputed, num.trees = 50,
                               probability = TRUE)
explainer_ranger  <- explain(titanic_ranger_model, data = titanic_imputed)
expl_ranger <- model_profile(explainer_ranger)
plot(expl_ranger)
plot(expl_ranger, geom = "aggregates")

vp_ra <- model_profile(explainer_ranger, type = "partial", variables = c("age", "fare"))
plot(vp_ra, variables = c("age", "fare"), geom = "points")

vp_ra <- model_profile(explainer_ranger, type = "partial", k = 3)
plot(vp_ra)
plot(vp_ra, geom = "profiles")
plot(vp_ra, geom = "points")

vp_ra <- model_profile(explainer_ranger, type = "partial", groups = "gender")
plot(vp_ra)
plot(vp_ra, geom = "profiles")
plot(vp_ra, geom = "points")

vp_ra <- model_profile(explainer_ranger, type = "accumulated")
plot(vp_ra)
plot(vp_ra, geom = "profiles")
plot(vp_ra, geom = "points")

Description

Plot Instance Level Residual Diagnostics

Usage

## S3 method for class 'predict_diagnostics'
plot(x, ...)

Arguments

Value

an ggplot2 object of the class gg.

Examples

library("ranger")
titanic_glm_model <- ranger(survived ~ gender + age + class + fare + sibsp + parch,
                     data = titanic_imputed)
explainer_glm <- explain(titanic_glm_model,
                         data = titanic_imputed,
                         y = titanic_imputed$survived)
johny_d <- titanic_imputed[24, c("gender", "age", "class", "fare", "sibsp", "parch")]

pl <- predict_diagnostics(explainer_glm, johny_d, variables = NULL)
plot(pl)

pl <- predict_diagnostics(explainer_glm, johny_d,
                       neighbors = 10,
                       variables = c("age", "fare"))
plot(pl)

pl <- predict_diagnostics(explainer_glm,
                       johny_d,
                       neighbors = 10,
                       variables = c("class", "gender"))
plot(pl)

Description

Plot Variable Attribution Explanations

Usage

## S3 method for class 'predict_parts'
plot(x, ...)

Arguments

Value

An object of the class ggplot.

Plot options

break_down

• max_features maximal number of features to be included in the plot. default value is 10

• min_max a range of OX axis. By default NA, therefore it will be extracted from the contributions of x. But it can be set to some constants, useful if these plots are to be used for comparisons.

• add_contributions if TRUE, variable contributions will be added to the plot.

• shift_contributions number describing how much labels should be shifted to the right, as a fraction of range. By default equal to 0.05.

• vcolors If NA (default), DrWhy colors are used.

• vnames a character vector, if specified then will be used as labels on OY axis. By default NULL.

• digits number of decimal places (round) or significant digits (signif) to be used.

• rounding_function a function to be used for rounding numbers.

• plot_distributions if TRUE then distributions of conditional propotions will be plotted. This requires keep_distributions=TRUE in the break_down, local_attributions, or local_interactions.

• baseline if numeric then veritical line starts in baseline.

• title a character. Plot title. By default "Break Down profile".

• subtitle a character. Plot subtitle. By default NULL - then subtitle is set to "created for the XXX, YYY model", where XXX, YYY are labels of given explainers.

• max_vars alias for the max_features parameter.

shap

• show_boxplots logical if TRUE (default) boxplot will be plotted to show uncertanity of attributions.

• vcolors If NA (default), DrWhy colors are used.

• max_features maximal number of features to be included in the plot. default value is 10

• max_vars alias for the max_features parameter.

oscillations

• bar_width width of bars. By default 10

Description

Plot Variable Profile Explanations

Usage

## S3 method for class 'predict_profile'
plot(x, ...)

Arguments

Value

An object of the class ggplot.

Plot options

ceteris_paribus

• color a character. Either name of a color or name of a variable that should be used for coloring

• size a numeric. Size of lines to be plotted

• alpha a numeric between 0 and 1. Opacity of lines

• facet_ncol number of columns for the facet_wrap

• variables if not NULL then only variables will be presented

• variable_type a character. If numerical then only numerical variables will be plotted. If categorical then only categorical variables will be plotted.

• title a character. Plot title. By default "Ceteris Paribus profile".

• subtitle a character. Plot subtitle. By default NULL - then subtitle is set to "created for the XXX, YYY model", where XXX, YYY are labels of given explainers.

• categorical_type a character. How categorical variables shall be plotted? Either "lines" (default) or "bars".

Description

Displays a waterfall aggregated shap plot for objects of shap_aggregated class.

Usage

## S3 method for class 'shap_aggregated'
plot(
  x,
  ...,
  shift_contributions = 0.05,
  add_contributions = TRUE,
  add_boxplots = TRUE,
  max_features = 10,
  title = "Aggregated SHAP"
)

Arguments

Value

a ggplot2 object.

Examples

library("DALEX")
set.seed(1313)
model_titanic_glm <- glm(survived ~ gender + age + fare,
                       data = titanic_imputed, family = "binomial")
explain_titanic_glm <- explain(model_titanic_glm,
                           data = titanic_imputed,
                           y = titanic_imputed$survived,
                           label = "glm")


bd_glm <- shap_aggregated(explain_titanic_glm, titanic_imputed[1:10, ])
bd_glm
plot(bd_glm)
plot(bd_glm, max_features = 3)
plot(bd_glm, max_features = 3,
     vnames = c("average","+ male","+ young","+ cheap ticket", "+ other factors", "final"))

Description

This function performs local diagnostic of residuals. For a single instance its neighbors are identified in the validation data. Residuals are calculated for neighbors and plotted against residuals for all data. Find information how to use this function here: https://ema.drwhy.ai/localDiagnostics.html.

Usage

predict_diagnostics(
  explainer,
  new_observation,
  variables = NULL,
  ...,
  nbins = 20,
  neighbors = 50,
  distance = gower::gower_dist
)

individual_diagnostics(
  explainer,
  new_observation,
  variables = NULL,
  ...,
  nbins = 20,
  neighbors = 50,
  distance = gower::gower_dist
)

Arguments

Value

An object of the class 'predict_diagnostics'. It's a data frame with calculated distribution of residuals.

References

Explanatory Model Analysis. Explore, Explain, and Examine Predictive Models. https://ema.drwhy.ai/

Examples

library("ranger")
titanic_glm_model <- ranger(survived ~ gender + age + class + fare + sibsp + parch,
                     data = titanic_imputed)
explainer_glm <- explain(titanic_glm_model,
                         data = titanic_imputed,
                         y = titanic_imputed$survived)
johny_d <- titanic_imputed[24, c("gender", "age", "class", "fare", "sibsp", "parch")]

id_johny <- predict_diagnostics(explainer_glm, johny_d, variables = NULL)
id_johny
plot(id_johny)

id_johny <- predict_diagnostics(explainer_glm, johny_d,
                       neighbors = 10,
                       variables = c("age", "fare"))
id_johny
plot(id_johny)

id_johny <- predict_diagnostics(explainer_glm,
                       johny_d,
                       neighbors = 10,
                       variables = c("class", "gender"))
id_johny
plot(id_johny)

Description

Instance Level Variable Attributions as Break Down, SHAP, aggregated SHAP or Oscillations explanations. Model prediction is decomposed into parts that are attributed for particular variables. From DALEX version 1.0 this function calls the break_down or shap functions from the iBreakDown package or ceteris_paribus from the ingredients package or kernelshap from the kernelshap package. Find information how to use the break_down method here: https://ema.drwhy.ai/breakDown.html. Find information how to use the shap method here: https://ema.drwhy.ai/shapley.html. Find information how to use the oscillations method here: https://ema.drwhy.ai/ceterisParibusOscillations.html. Find information how to use the kernelshap method here: https://modeloriented.github.io/kernelshap/ aSHAP method provides explanations for a set of observations based on SHAP.

Usage

predict_parts(
  explainer,
  new_observation,
  ...,
  N = if (substr(type, 1, 4) == "osci") 500 else NULL,
  type = "break_down"
)

predict_parts_oscillations(explainer, new_observation, ...)

predict_parts_oscillations_uni(
  explainer,
  new_observation,
  variable_splits_type = "uniform",
  ...
)

predict_parts_oscillations_emp(
  explainer,
  new_observation,
  variable_splits = NULL,
  variables = colnames(explainer$data),
  ...
)

predict_parts_break_down(explainer, new_observation, ...)

predict_parts_break_down_interactions(explainer, new_observation, ...)

predict_parts_shap(explainer, new_observation, ...)

predict_parts_shap_aggregated(explainer, new_observation, ...)

predict_parts_kernel_shap(explainer, new_observation, ...)

predict_parts_kernel_shap_break_down(explainer, new_observation, ...)

predict_parts_kernel_shap_aggreagted(explainer, new_observation, ...)

variable_attribution(
  explainer,
  new_observation,
  ...,
  N = if (substr(type, 1, 4) == "osci") 500 else NULL,
  type = "break_down"
)

Arguments

Value

Depending on the type there are different classes of the resulting object. It's a data frame with calculated average response.

References

Explanatory Model Analysis. Explore, Explain, and Examine Predictive Models. https://ema.drwhy.ai/

Examples

library(DALEX)

new_dragon <- data.frame(
    year_of_birth = 200,
    height = 80,
    weight = 12.5,
    scars = 0,
    number_of_lost_teeth  = 5
)

model_lm <- lm(life_length ~ year_of_birth + height +
               weight + scars + number_of_lost_teeth,
               data = dragons)

explainer_lm <- explain(model_lm,
                        data = dragons,
                        y = dragons$year_of_birth,
                        label = "model_lm")

bd_lm <- predict_parts_break_down(explainer_lm, new_observation = new_dragon)
head(bd_lm)
plot(bd_lm)


library("ranger")
model_ranger <- ranger(life_length ~ year_of_birth + height +
                       weight + scars + number_of_lost_teeth,
                       data = dragons, num.trees = 50)

explainer_ranger <- explain(model_ranger,
                            data = dragons,
                            y = dragons$year_of_birth,
                            label = "model_ranger")

bd_ranger <- predict_parts_break_down(explainer_ranger, new_observation = new_dragon)
head(bd_ranger)
plot(bd_ranger)

Description

This function calculated individual profiles aka Ceteris Paribus Profiles. From DALEX version 1.0 this function calls the ceteris_paribus from the ingredients package. Find information how to use this function here: https://ema.drwhy.ai/ceterisParibus.html.

Usage

predict_profile(
  explainer,
  new_observation,
  variables = NULL,
  ...,
  type = "ceteris_paribus",
  variable_splits_type = "uniform"
)

individual_profile(
  explainer,
  new_observation,
  variables = NULL,
  ...,
  type = "ceteris_paribus",
  variable_splits_type = "uniform"
)

Arguments

Value

An object of the class ceteris_paribus_explainer. It's a data frame with calculated average response.

References

Explanatory Model Analysis. Explore, Explain, and Examine Predictive Models. https://ema.drwhy.ai/

Examples

new_dragon <- data.frame(year_of_birth = 200,
     height = 80,
     weight = 12.5,
     scars = 0,
     number_of_lost_teeth  = 5)

dragon_lm_model4 <- lm(life_length ~ year_of_birth + height +
                                     weight + scars + number_of_lost_teeth,
                       data = dragons)
dragon_lm_explainer4 <- explain(dragon_lm_model4, data = dragons, y = dragons$year_of_birth,
                                label = "model_4v")
dragon_lm_predict4 <- predict_profile(dragon_lm_explainer4,
                new_observation = new_dragon,
                variables = c("year_of_birth", "height", "scars"))
head(dragon_lm_predict4)
plot(dragon_lm_predict4,
    variables = c("year_of_birth", "height", "scars"))


library("ranger")
dragon_ranger_model4 <- ranger(life_length ~ year_of_birth + height +
                                               weight + scars + number_of_lost_teeth,
                                 data = dragons, num.trees = 50)
dragon_ranger_explainer4 <- explain(dragon_ranger_model4, data = dragons, y = dragons$year_of_birth,
                                label = "model_ranger")
dragon_ranger_predict4 <- predict_profile(dragon_ranger_explainer4,
                                           new_observation = new_dragon,
                                           variables = c("year_of_birth", "height", "scars"))
head(dragon_ranger_predict4)
plot(dragon_ranger_predict4,
    variables = c("year_of_birth", "height", "scars"))

Description

This is a generic predict() function works for explainer objects.

Usage

## S3 method for class 'explainer'
predict(object, newdata, ...)

model_prediction(explainer, new_data, ...)

Arguments

Value

An numeric matrix of predictions

Examples

HR_glm_model <- glm(status == "fired"~., data = HR, family = "binomial")
explainer_glm <- explain(HR_glm_model, data = HR)
predict(explainer_glm, HR[1:3,])

 
library("ranger")
HR_ranger_model <- ranger(status~., data = HR, num.trees = 50, probability = TRUE)
explainer_ranger  <- explain(HR_ranger_model, data = HR)
predict(explainer_ranger, HR[1:3,])

model_prediction(explainer_ranger, HR[1:3,])

Description

Generic function

Usage

## S3 method for class 'description'
print(x, ...)

Arguments

Description

Print Explainer Summary

Usage

## S3 method for class 'explainer'
print(x, ...)

Arguments

Examples

aps_lm_model4 <- lm(m2.price~., data = apartments)
aps_lm_explainer4 <- explain(aps_lm_model4, data = apartments, y = apartments$m2.price,
                             label = "model_4v")
aps_lm_explainer4

 
library("ranger")
titanic_ranger_model <- ranger(survived~., data = titanic_imputed, num.trees = 50,
                               probability = TRUE)
explainer_ranger  <- explain(titanic_ranger_model, data = titanic_imputed[,-8],
                             y = titanic_imputed$survived,
                             label = "model_ranger")
explainer_ranger

Description

Generic function

Usage

## S3 method for class 'model_diagnostics'
print(x, ...)

Arguments

Description

Function prints object of class model_info created with model_info

Usage

## S3 method for class 'model_info'
print(x, ...)

Arguments

Description

Print Dataset Level Model Performance Summary

Usage

## S3 method for class 'model_performance'
print(x, ...)

Arguments

Examples

library("ranger")
titanic_ranger_model <- ranger(survived~., data = titanic_imputed, num.trees = 100,
                               probability = TRUE)
# It's a good practice to pass data without target variable
explainer_ranger  <- explain(titanic_ranger_model, data = titanic_imputed[,-8],
                             y = titanic_imputed$survived)
# resulting dataframe has predicted values and residuals
mp_ex_rn <- model_performance(explainer_ranger)
mp_ex_rn
plot(mp_ex_rn)

Description

Generic function

Usage

## S3 method for class 'model_profile'
print(x, ...)

Arguments

Description

Generic function

Usage

## S3 method for class 'predict_diagnostics'
print(x, ...)

Arguments

Description

Default Theme for DALEX plots

Usage

set_theme_dalex(
  default_theme = "drwhy",
  default_theme_vertical = default_theme
)

theme_default_dalex()

theme_vertical_default_dalex()

Arguments

Value

list with current default themes

Examples

old <- set_theme_dalex("ema")

library("ranger")
apartments_ranger_model <- ranger(m2.price~., data = apartments, num.trees = 50)
explainer_ranger  <- explain(apartments_ranger_model, data = apartments[,-1],
                             y = apartments$m2.price, label = "Ranger Apartments")
model_parts_ranger_aps <- model_parts(explainer_ranger, type = "raw")
head(model_parts_ranger_aps, 8)
plot(model_parts_ranger_aps)

old <- set_theme_dalex(ggplot2::theme_void(), ggplot2::theme_void())
plot(model_parts_ranger_aps)

old <- set_theme_dalex("drwhy")
plot(model_parts_ranger_aps)
old <- set_theme_dalex(ggplot2::theme_void(), ggplot2::theme_void())
plot(model_parts_ranger_aps)

Description

This function works in a similar way to shap function from iBreakDown but it calculates explanations for a set of observation and then aggregates them.

Usage

shap_aggregated(
  explainer,
  new_observations,
  order = NULL,
  B = 25,
  kernelshap = FALSE,
  ...
)

Arguments

Value

an object of the shap_aggregated class.

References

Explanatory Model Analysis. Explore, Explain and Examine Predictive Models. https://ema.drwhy.ai

Examples

library("DALEX")
set.seed(1313)
model_titanic_glm <- glm(survived ~ gender + age + fare,
                       data = titanic_imputed, family = "binomial")
explain_titanic_glm <- explain(model_titanic_glm,
                           data = titanic_imputed,
                           y = titanic_imputed$survived,
                           label = "glm")


bd_glm <- shap_aggregated(explain_titanic_glm, titanic_imputed[1:10, ])
bd_glm
plot(bd_glm, max_features = 3)

Description

DrWhy Theme for ggplot objects

Usage

theme_drwhy()

theme_ema()

theme_drwhy_vertical()

theme_ema_vertical()

Value

theme for ggplot2 objects

Description

The titanic data is a complete list of passengers and crew members on the RMS Titanic. It includes a variable indicating whether a person did survive the sinking of the RMS Titanic on April 15, 1912.

Usage

data(titanic)
data(titanic_imputed)

Format

a data frame with 2207 rows and 9 columns

Details

This dataset was copied from the stablelearner package and went through few variable transformations. Levels in embarked was replaced with full names, sibsp, parch and fare were converted to numerical variables and values for crew were replaced with 0. If you use this dataset please cite the original package.

From stablelearner: The website https://www.encyclopedia-titanica.org offers detailed information about passengers and crew members on the RMS Titanic. According to the website 1317 passengers and 890 crew member were abord. 8 musicians and 9 employees of the shipyard company are listed as passengers, but travelled with a free ticket, which is why they have NA values in fare. In addition to that, fare is truely missing for a few regular passengers.

• gender a factor with levels male and female.

• age a numeric value with the persons age on the day of the sinking.

• class a factor specifying the class for passengers or the type of service aboard for crew members.

• embarked a factor with the persons place of of embarkment (Belfast/Cherbourg/Queenstown/Southampton).

• country a factor with the persons home country.

• fare a numeric value with the ticket price (0 for crew members, musicians and employees of the shipyard company).

• sibsp an ordered factor specifying the number if siblings/spouses aboard; adopted from Vanderbild data set (see below).

• parch an ordered factor specifying the number of parents/children aboard; adopted from Vanderbild data set (see below).

• survived a factor with two levels (no and yes) specifying whether the person has survived the sinking.

NOTE: The titanic_imputed dataset use following imputation rules.

• Missing 'age' is replaced with the mean of the observed ones, i.e., 30.

• For sibsp and parch, missing values are replaced by the most frequently observed value, i.e., 0.

• For fare, mean fare for a given class is used, i.e., 0 pounds for crew, 89 pounds for the 1st, 22 pounds for the 2nd, and 13 pounds for the 3rd class.

Source

This dataset was copied from the stablelearner package and went through few variable transformations. The complete list of persons on the RMS titanic was downloaded from https://www.encyclopedia-titanica.org on April 5, 2016. The information given in sibsp and parch was adopoted from a data set obtained from https://biostat.app.vumc.org/wiki/Main/DataSets.

References

https://www.encyclopedia-titanica.org and https://CRAN.R-project.org/package=stablelearner

Description

Function allows users to update data an y of any explainer in a unified way. It doesn't require knowledge about structre of an explainer.

Usage

update_data(explainer, data, y = NULL, verbose = TRUE)

Arguments

Value

updated explainer object

Examples

aps_lm_model4 <- lm(m2.price ~., data = apartments)
aps_lm_explainer4 <- explain(aps_lm_model4, data = apartments, label = "model_4v")
explainer <- update_data(aps_lm_explainer4, data = apartmentsTest, y = apartmentsTest$m2.price)

Description

Function allows users to update label of any explainer in a unified way. It doesn't require knowledge about structre of an explainer.

Usage

update_label(explainer, label, verbose = TRUE)

Arguments

Value

updated explainer object

Examples

aps_lm_model4 <- lm(m2.price ~., data = apartments)
aps_lm_explainer4 <- explain(aps_lm_model4, data = apartments, label = "model_4v")
explainer <- update_label(aps_lm_explainer4, label = "lm")

Description

From DALEX version 1.0 this function calls the accumulated_dependence or partial_dependence from the ingredients package. Find information how to use this function here: https://ema.drwhy.ai/partialDependenceProfiles.html.

Usage

variable_effect(explainer, variables, ..., type = "partial_dependency")

variable_effect_partial_dependency(explainer, variables, ...)

variable_effect_accumulated_dependency(explainer, variables, ...)

Arguments

Value

An object of the class 'aggregated_profiles_explainer'. It's a data frame with calculated average response.

References

Explanatory Model Analysis. Explore, Explain, and Examine Predictive Models. https://ema.drwhy.ai/

Examples

titanic_glm_model <- glm(survived~., data = titanic_imputed, family = "binomial")
explainer_glm <- explain(titanic_glm_model, data = titanic_imputed)
expl_glm <- variable_effect(explainer_glm, "fare", "partial_dependency")
plot(expl_glm)

 
library("ranger")
titanic_ranger_model <- ranger(survived~., data = titanic_imputed, num.trees = 50,
                               probability = TRUE)
explainer_ranger  <- explain(titanic_ranger_model, data = titanic_imputed)
expl_ranger  <- variable_effect(explainer_ranger, variables = "fare",
                            type = "partial_dependency")
plot(expl_ranger)
plot(expl_ranger, expl_glm)

# Example for factor variable (with factorMerger)
expl_ranger_factor  <- variable_effect(explainer_ranger, variables =  "class")
plot(expl_ranger_factor)

Description

This function is a wrapper over various predict functions for different models and differnt model structures. The wrapper returns a single numeric score for each new observation. To do this it uses different extraction techniques for models from different classes, like for classification random forest is forces the output to be probabilities not classes itself.

Usage

yhat(X.model, newdata, ...)

## S3 method for class 'lm'
yhat(X.model, newdata, ...)

## S3 method for class 'randomForest'
yhat(X.model, newdata, ...)

## S3 method for class 'svm'
yhat(X.model, newdata, ...)

## S3 method for class 'gbm'
yhat(X.model, newdata, ...)

## S3 method for class 'glm'
yhat(X.model, newdata, ...)

## S3 method for class 'cv.glmnet'
yhat(X.model, newdata, ...)

## S3 method for class 'glmnet'
yhat(X.model, newdata, ...)

## S3 method for class 'ranger'
yhat(X.model, newdata, ...)

## S3 method for class 'model_fit'
yhat(X.model, newdata, ...)

## S3 method for class 'train'
yhat(X.model, newdata, ...)

## S3 method for class 'lrm'
yhat(X.model, newdata, ...)

## S3 method for class 'rpart'
yhat(X.model, newdata, ...)

## S3 method for class ''function''
yhat(X.model, newdata, ...)

## S3 method for class 'party'
yhat(X.model, newdata, ...)

## Default S3 method:
yhat(X.model, newdata, ...)

Arguments

Details

Currently supported packages are:

• class cv.glmnet and glmnet - models created with glmnet package,

• class glm - generalized linear models created with glm,

• class model_fit - models created with parsnip package,

• class lm - linear models created with lm,

• class ranger - models created with ranger package,

• class randomForest - random forest models created with randomForest package,

• class svm - support vector machines models created with the e1071 package,

• class train - models created with caret package,

• class gbm - models created with gbm package,

• class lrm - models created with rms package,

• class rpart - models created with rpart package.

Value

An numeric matrix of predictions