Standardised Statistical Comparison Workflows

---

Documentation for package ‘STATassist’ version 1.0.0

• DESCRIPTION file.
• Package NEWS.

Help Pages

as.table.sa_categorical	The contingency table a categorical comparison was run on
center_by_control	Centre every feature on the control group
cluster_dbscan	Cluster by finding the dense regions
cluster_hclust	Cluster by building a tree and cutting it
cluster_kmeans	Cluster by moving centres until they stop
cluster_snn	Cluster by how many neighbours points have in common
coef.sa_fit	Coefficients, summary and predictions from the model inside a fit
coef.sa_model	Coefficients of a fitted model
compare_categorical_groups	Test a contingency table with every applicable test at once
compare_factorial_groups	Analyse a crossed-factor design as one model
compare_multiple_groups	Run every applicable multi-group test at once
compare_one_sample	Compare one sample against a hypothesised value
compare_two_groups	Run every applicable two-group test at once
diagnose_distribution	Check the assumptions a comparison rests on
draw_butterfly_hist	Draw a butterfly histogram of one feature across two groups
draw_corrplot	Draw a correlation matrix, with the cells that failed the test left blank
draw_dim_reduction_plot	Draw a reduction as a scatter of its points
draw_forest_plot	Draw a forest plot of a comparison result
draw_grouped_barplot	Draw a grouped barplot of a descriptive summary
draw_grouped_boxplot	Draw a grouped boxplot across several features
draw_heatmap	Draw a clustered heatmap of features by samples
draw_interaction_plot	Draw an interaction plot of a factorial comparison
draw_mosaic_plot	Draw a mosaic plot of a contingency table
draw_prediction_plot	Draw predicted against observed for an evaluated regression
draw_roc_curve	Draw the ROC curves of an evaluated classification
draw_volcano_plot	Draw a volcano plot
estimate_categorical_significance	Reduce a contingency table to significance verdicts
estimate_significance	Reduce a comparison to one significance verdict per feature
evaluate_classification_models	Score fitted classifications on held-out rows
evaluate_regression_models	Score fitted regressions on held-out rows
fit_elastic_net	Fit an elastic net, lasso or ridge regression
fit_linear_regression	Fit a linear regression
fit_logistic_regression	Fit a logistic regression
fit_rf	Fit a random forest
fit_svm	Fit a support vector machine
make_block_cor	Build a block correlation matrix
perform_pca	Reduce many features to a few components
perform_rfe	Select the predictors worth keeping
perform_stepwise	Stepwise feature selection by information criteria
perform_tsne	Embed samples or features with t-SNE
perform_umap	Embed samples or features with UMAP
plot.sa_categorical	Draw a mosaic plot of a contingency table
plot.sa_comparison	Draw a forest plot of a comparison result
plot.sa_performance	Draw an evaluation result
plot.sa_reduction	Draw a reduction as a scatter of its points
predict.sa_fit	Coefficients, summary and predictions from the model inside a fit
predict.sa_model	Predict from a fitted model on rows it was not fitted to
print.sa_categorical	Print a categorical comparison
print.sa_categorical_significance	Print a categorical significance verdict
print.sa_cluster	Print a clustering
print.sa_comparison	Print a comparison result
print.sa_diagnosis	Print a distribution diagnosis
print.sa_model	Print a fitted model
print.sa_performance	Print an evaluation result
print.sa_reduction	Print a dimensionality reduction
print.sa_selection	Print a feature selection
print.sa_significance	Print a significance verdict
print.sa_split	Print a train/test split
screen_outliers	Flag candidate outliers without removing them
simulate_categorical_groups	Simulate a contingency table whose association is known
simulate_classification	Simulate a two-class outcome whose coefficients are known
simulate_factorial_groups	Simulate a crossed-factor experiment whose answer is known
simulate_multiple_groups	Simulate a control-versus-treatments experiment whose answer is known
simulate_regression	Simulate a regression whose coefficients are known
simulate_two_groups	Simulate a two-group experiment whose answer is known
split_data	Split data into training and test sets
summarize_association_stats	Correlation between every pair of features, with all three coefficients
summarize_descriptive_stats	Descriptive summary of several features, optionally split by group
summary.sa_fit	Coefficients, summary and predictions from the model inside a fit