PlotMiningMatrix

Generate a matrix of statistical relationships between variables.

Usage

PlotMiningMatrix(
  data,
  outcome_vars,
  predictor_vars = NULL,
  covariates = NULL,
  Relabel = TRUE,
  TreatOrdinalAs = "Categorical",
  Parametric = TRUE,
  Data = lifecycle::deprecated(),
  OutcomeVars = lifecycle::deprecated(),
  PredictorVars = lifecycle::deprecated(),
  fdr_scope = c("matrix", "per_outcome", "per_predictor"),
  Covariates = lifecycle::deprecated()
)

Arguments

data

A data frame.

outcome_vars

Outcome variables.

predictor_vars

Predictor variables. If NULL, uses OutcomeVars.

covariates

Optional covariates (reserved for future use).

Relabel

Use labels instead of names.

TreatOrdinalAs

How ordinal variables are handled: "Categorical", "Continuous", "Both", or "Exclude".

Parametric

Use parametric tests.

Data

Deprecated (since 19.15.0). Use data instead.

OutcomeVars

Deprecated (since 19.15.0). Use outcome_vars instead.

PredictorVars

Deprecated (since 19.15.0). Use predictor_vars instead.

fdr_scope

Either "matrix" (default) or "per_outcome", passed to ApplyFDRCorrection(). "matrix" corrects across all pairwise p-values at once (historical behavior, computed on the symmetrized pair table). "per_outcome" corrects separately within each x-axis variable (XVar, ordered by outcome_vars).

Covariates

Deprecated (since 19.15.0). Use covariates instead.

Value

List with tables and plots.

Examples

data(SampleData)
data(SampleVariableTypes)

# Attach labels and factor levels for readable axes
Labelled <- RevalueData(SampleData, SampleVariableTypes)$RevaluedData

# A mining matrix over 11 mixed-type variables (categorical + continuous)
result <- PlotMiningMatrix(
  Labelled,
  outcome_vars   = c("Diagnosis", "sex", "age", "AXL", "Adiponectin"),
  predictor_vars = c("Alpha_1_Antitrypsin", "Alpha_2_Macroglobulin",
                     "Apolipoprotein_B", "C_Reactive_Protein",
                     "Cortisol", "Insulin")
)

# The relationship plot (point shape/size encode significance from raw p)
result$Unadjusted$plot


# The p-value table carries both unadjusted (p) and FDR-adjusted (p_adj)
# p-values so results can be inspected with and without FDR correction
result$Unadjusted$PvalTable[, c("XVar", "YVar", "p", "p_adj", "Test")]
#> # A tibble: 18 × 5
#>    XVar        YVar                         p    p_adj Test       
#>    <chr>       <chr>                    <dbl>    <dbl> <chr>      
#>  1 AXL         Alpha_1_Antitrypsin   1.12e- 1 2.01e- 1 Correlation
#>  2 AXL         Alpha_2_Macroglobulin 8.35e-15 5.01e-14 Correlation
#>  3 AXL         Apolipoprotein_B      6.02e- 1 6.78e- 1 Correlation
#>  4 AXL         C_Reactive_Protein    4.82e- 1 6.58e- 1 Correlation
#>  5 AXL         Cortisol              1.85e- 2 4.15e- 2 Correlation
#>  6 AXL         Insulin               2.29e-14 1.03e-13 Correlation
#>  7 Adiponectin Alpha_1_Antitrypsin   1.21e-15 1.09e-14 Correlation
#>  8 Adiponectin Alpha_2_Macroglobulin 3.24e-12 1.17e-11 Correlation
#>  9 Adiponectin Apolipoprotein_B      2.00e-18 3.60e-17 Correlation
#> 10 Adiponectin C_Reactive_Protein    1.02e- 1 2.01e- 1 Correlation
#> 11 Adiponectin Cortisol              9.14e- 3 2.35e- 2 Correlation
#> 12 Adiponectin Insulin               7.08e- 4 2.12e- 3 Correlation
#> 13 age         Alpha_1_Antitrypsin   7.17e- 1 7.60e- 1 Correlation
#> 14 age         Alpha_2_Macroglobulin 3.58e- 1 5.36e- 1 Correlation
#> 15 age         Apolipoprotein_B      5.25e- 1 6.58e- 1 Correlation
#> 16 age         C_Reactive_Protein    3.49e- 1 5.36e- 1 Correlation
#> 17 age         Cortisol              9.60e- 1 9.60e- 1 Correlation
#> 18 age         Insulin               5.48e- 1 6.58e- 1 Correlation

# Per-outcome FDR correction instead of matrix-wide
result_perout <- PlotMiningMatrix(
  Labelled,
  outcome_vars   = c("Diagnosis", "sex", "age", "AXL", "Adiponectin"),
  predictor_vars = c("Alpha_1_Antitrypsin", "Alpha_2_Macroglobulin",
                     "Apolipoprotein_B", "C_Reactive_Protein",
                     "Cortisol", "Insulin"),
  fdr_scope = "per_outcome"
)

# An interactive version of the matrix
if (requireNamespace("plotly", quietly = TRUE)) {
  plotly::ggplotly(result$Unadjusted$plot)
}