SHAP values are used to cluster data samples using the k-means method to identify subgroups of individuals with specific patterns of feature contributions.
Usage
SHAPclust(
task,
trained_model,
splits,
shap_Mean_wide,
shap_Mean_long,
num_of_clusters = 4,
seed = 246,
subset = 1,
algorithm = "Hartigan-Wong",
iter.max = 1000
)Arguments
- task
an mlr3 task for binary classification
- trained_model
an mlr3 trained learner object
- splits
an mlr3 object defining data splits for train and test sets
- shap_Mean_wide
the data frame of SHAP values in wide format from eSHAP_plot.R
- shap_Mean_long
the data frame of SHAP values in long format from eSHAP_plot.R
- num_of_clusters
number of clusters to make based on SHAP values, default: 4
- seed
an integer for reproducibility, Default to 246
- subset
what percentage of the instances to use from 0 to 1 where 1 means all
- algorithm
k-means algorithm character: "Hartigan-Wong", "Lloyd", "Forgy", "MacQueen".
- iter.max
maximum number of iterations allowed
Value
A list containing four elements:
- shap_plot_onerow
An interactive plot displaying the SHAP values for each feature, clustered by the specified number of clusters. Each cluster is shown in a facet.
- combined_plot
A ggplot2 figure combining confusion matrices for each cluster, providing insights into the model's performance within each identified subgroup.
- kmeans_fvals_desc
A summary table containing statistical descriptions of the clusters based on feature values.
- shap_Mean_wide_kmeans
A data frame containing clustered SHAP values along with predictions and ground truth information.
- kmeans_info
Information about the k-means clustering process, including cluster centers and assignment details.
References
Zargari Marandi, R., 2024. ExplaineR: an R package to explain machine learning models. Bioinformatics advances, 4(1), p.vbae049, https://doi.org/10.1093/bioadv/vbae049.
See also
Other functions to visualize and interpret machine learning models: eSHAP_plot.
Examples
# \donttest{
library("explainer")
seed <- 246
set.seed(seed)
# Load necessary packages
if (!requireNamespace("mlbench", quietly = TRUE)) stop("mlbench not installed.")
if (!requireNamespace("mlr3learners", quietly = TRUE)) stop("mlr3learners not installed.")
if (!requireNamespace("ranger", quietly = TRUE)) stop("ranger not installed.")
# Load BreastCancer dataset
utils::data("BreastCancer", package = "mlbench")
target_col <- "Class"
positive_class <- "malignant"
mydata <- BreastCancer[, -1]
mydata <- na.omit(mydata)
sex <- sample(
c("Male", "Female"),
size = nrow(mydata),
replace = TRUE
)
mydata$age <- as.numeric(sample(
seq(18, 60),
size = nrow(mydata),
replace = TRUE
))
mydata$sex <- factor(
sex,
levels = c("Male", "Female"),
labels = c(1, 0)
)
maintask <- mlr3::TaskClassif$new(
id = "my_classification_task",
backend = mydata,
target = target_col,
positive = positive_class
)
splits <- mlr3::partition(maintask)
mylrn <- mlr3::lrn(
"classif.ranger",
predict_type = "prob"
)
mylrn$train(maintask, splits$train)
SHAP_output <- eSHAP_plot(
task = maintask,
trained_model = mylrn,
splits = splits,
sample.size = 2, # also 30 or more
seed = seed,
subset = 0.02 # up to 1
)
#> Warning: Ignoring unknown aesthetics: text
shap_Mean_wide <- SHAP_output[[2]]
shap_Mean_long <- SHAP_output[[3]]
SHAP_plot_clusters <- SHAPclust(
task = maintask,
trained_model = mylrn,
splits = splits,
shap_Mean_wide = shap_Mean_wide,
shap_Mean_long = shap_Mean_long,
num_of_clusters = 3, # your choice
seed = seed,
subset = 0.02, # match with eSHAP_plot
algorithm = "Hartigan-Wong",
iter.max = 10
)
#> Key: <sample_num, feature, Phi>
#> sample_num feature Phi cluster mean_phi f_val
#> <int> <char> <num> <int> <num> <num>
#> 1: 1 Bare.nuclei -0.1399936508 2 0.0493383117 0.0000000
#> 2: 1 Bl.cromatin -0.0828492063 2 0.0489450397 0.0000000
#> 3: 1 Cell.shape -0.0367019841 2 0.0745567280 0.0000000
#> 4: 1 Cell.size 0.0000000000 2 0.0319849206 0.0000000
#> 5: 1 Cl.thickness -0.0178948413 2 0.0262465278 0.5000000
#> 6: 1 Epith.c.size -0.0055142857 2 0.0059076389 0.5000000
#> 7: 1 Marg.adhesion 0.0000000000 2 0.0238763889 0.0000000
#> 8: 1 Mitoses 0.0000000000 2 0.0043198413 0.0000000
#> 9: 1 Normal.nucleoli -0.0798761905 2 0.0537973214 0.0000000
#> 10: 1 age -0.0151575397 2 0.0140473214 0.9444444
#> 11: 1 sex -0.0010932540 2 0.0005277778 NaN
#> 12: 2 Bare.nuclei -0.0050662698 3 0.0493383117 1.0000000
#> 13: 2 Bl.cromatin 0.0610115079 3 0.0489450397 1.0000000
#> 14: 2 Cell.shape 0.1521496032 3 0.0745567280 1.0000000
#> 15: 2 Cell.size 0.1072460317 3 0.0319849206 1.0000000
#> 16: 2 Cl.thickness 0.0501396825 3 0.0262465278 1.0000000
#> 17: 2 Epith.c.size 0.0181162698 3 0.0059076389 1.0000000
#> 18: 2 Marg.adhesion 0.0858515873 3 0.0238763889 1.0000000
#> 19: 2 Mitoses 0.0172793651 3 0.0043198413 1.0000000
#> 20: 2 Normal.nucleoli 0.0572436508 3 0.0537973214 1.0000000
#> 21: 2 age -0.0293571429 3 0.0140473214 1.0000000
#> 22: 2 sex 0.0010178571 3 0.0005277778 NaN
#> 23: 3 Bare.nuclei 0.0000000000 1 0.0493383117 0.0000000
#> 24: 3 Bl.cromatin 0.0000000000 1 0.0489450397 0.5000000
#> 25: 3 Cell.shape 0.0000000000 1 0.0745567280 0.0000000
#> 26: 3 Cell.size 0.0000000000 1 0.0319849206 0.0000000
#> 27: 3 Cl.thickness 0.0002857143 1 0.0262465278 0.6666667
#> 28: 3 Epith.c.size 0.0000000000 1 0.0059076389 0.0000000
#> 29: 3 Marg.adhesion 0.0000000000 1 0.0238763889 0.0000000
#> 30: 3 Mitoses 0.0000000000 1 0.0043198413 0.0000000
#> 31: 3 Normal.nucleoli 0.0000000000 1 0.0537973214 0.0000000
#> 32: 3 age 0.0002857143 1 0.0140473214 0.0000000
#> 33: 3 sex 0.0000000000 1 0.0005277778 NaN
#> 34: 4 Bare.nuclei -0.0522933261 2 0.0493383117 0.0000000
#> 35: 4 Bl.cromatin -0.0519194444 2 0.0489450397 0.0000000
#> 36: 4 Cell.shape -0.1093753247 2 0.0745567280 0.0000000
#> 37: 4 Cell.size -0.0206936508 2 0.0319849206 0.0000000
#> 38: 4 Cl.thickness -0.0366658730 2 0.0262465278 0.0000000
#> 39: 4 Epith.c.size 0.0000000000 2 0.0059076389 0.0000000
#> 40: 4 Marg.adhesion -0.0096539683 2 0.0238763889 0.0000000
#> 41: 4 Mitoses 0.0000000000 2 0.0043198413 0.0000000
#> 42: 4 Normal.nucleoli -0.0780694444 2 0.0537973214 0.0000000
#> 43: 4 age -0.0113888889 2 0.0140473214 0.4722222
#> 44: 4 sex 0.0000000000 2 0.0005277778 NaN
#> sample_num feature Phi cluster mean_phi f_val
#> unscaled_f_val correct_prediction pred_prob pred_class
#> <num> <fctr> <num> <fctr>
#> 1: 1 Correct 0.0043809524 benign
#> 2: 1 Correct 0.0043809524 benign
#> 3: 1 Correct 0.0043809524 benign
#> 4: 1 Correct 0.0043809524 benign
#> 5: 4 Correct 0.0043809524 benign
#> 6: 3 Correct 0.0043809524 benign
#> 7: 1 Correct 0.0043809524 benign
#> 8: 1 Correct 0.0043809524 benign
#> 9: 1 Correct 0.0043809524 benign
#> 10: 54 Correct 0.0043809524 benign
#> 11: 2 Correct 0.0043809524 benign
#> 12: 3 Correct 0.9575134921 malignant
#> 13: 8 Correct 0.9575134921 malignant
#> 14: 7 Correct 0.9575134921 malignant
#> 15: 8 Correct 0.9575134921 malignant
#> 16: 7 Correct 0.9575134921 malignant
#> 17: 4 Correct 0.9575134921 malignant
#> 18: 6 Correct 0.9575134921 malignant
#> 19: 4 Correct 0.9575134921 malignant
#> 20: 8 Correct 0.9575134921 malignant
#> 21: 56 Correct 0.9575134921 malignant
#> 22: 2 Correct 0.9575134921 malignant
#> 23: 1 Correct 0.0005714286 benign
#> 24: 3 Correct 0.0005714286 benign
#> 25: 1 Correct 0.0005714286 benign
#> 26: 1 Correct 0.0005714286 benign
#> 27: 5 Correct 0.0005714286 benign
#> 28: 2 Correct 0.0005714286 benign
#> 29: 1 Correct 0.0005714286 benign
#> 30: 1 Correct 0.0005714286 benign
#> 31: 1 Correct 0.0005714286 benign
#> 32: 20 Correct 0.0005714286 benign
#> 33: 2 Correct 0.0005714286 benign
#> 34: 1 Correct 0.0000000000 benign
#> 35: 1 Correct 0.0000000000 benign
#> 36: 1 Correct 0.0000000000 benign
#> 37: 1 Correct 0.0000000000 benign
#> 38: 1 Correct 0.0000000000 benign
#> 39: 2 Correct 0.0000000000 benign
#> 40: 1 Correct 0.0000000000 benign
#> 41: 1 Correct 0.0000000000 benign
#> 42: 1 Correct 0.0000000000 benign
#> 43: 37 Correct 0.0000000000 benign
#> 44: 2 Correct 0.0000000000 benign
#> unscaled_f_val correct_prediction pred_prob pred_class
#> Warning: Ignoring unknown aesthetics: text
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: no non-missing arguments to max; returning -Inf
#> Warning: Computation failed in `stat_ydensity()`.
#> Caused by error in `$<-.data.frame`:
#> ! replacement has 1 row, data has 0
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: Groups with fewer than two datapoints have been dropped.
#> ℹ Set `drop = FALSE` to consider such groups for position adjustment purposes.
#> Warning: no non-missing arguments to max; returning -Inf
#> Warning: Computation failed in `stat_ydensity()`.
#> Caused by error in `$<-.data.frame`:
#> ! replacement has 1 row, data has 0
#> Warning: 'ggimage' is missing. Will not plot arrows and zero-shading.
#> Warning: 'rsvg' is missing. Will not plot arrows and zero-shading.
#> Warning: 'ggimage' is missing. Will not plot arrows and zero-shading.
#> Warning: 'rsvg' is missing. Will not plot arrows and zero-shading.
#> Warning: 'ggimage' is missing. Will not plot arrows and zero-shading.
#> Warning: 'rsvg' is missing. Will not plot arrows and zero-shading.
# }