Skip to contents

Group Comparison Table

Source: R/Table_GroupComparison.R
Table_GroupComparison.Rd

The statistics of the view the group comparison chart opens on, as a table: one row per panel the chart tests, with R's method, each estimate and its interval, the counts, the p-value and its note, written as the chart prints them. The numbers are Analyze_GroupDifference()'s on the rows the chart draws in each panel, as Widget_GroupComparison() stores them; with pairwise = TRUE each pair of groups has a row of its own beneath its panel.

Usage

Table_GroupComparison(dfResults, dfParticipants = NULL, lSettings = list())

Arguments

dfResults

data.frame Long-format results, one row per participant, biomarker and visit. Column names are supplied by lSettings; the defaults expect USUBJID/TEST/STRESN/VISIT/VISITNUM/STRESU, the columns of Synthetic_Results.

dfParticipants

data.frame One row per participant, or NULL. With it the chart has filters, and offers its category columns to group, colour and panel by; it also says who the participants are. Without it a group comes from a column carried on the results rows. Default: NULL.

lSettings

list bio.viz group comparison settings, as Widget_GroupComparison() takes them, and title, subtitle and footnotes. A table is of one biomarker, so start_value must name it. Default: list().

Value

A data.frame of text: Visit (and Panel with panel_by, and Pair with pairwise), then Statistic, Method, Estimate, Counts, p-value and Note. Its attributes are title, subtitle, footnotes, results (R's answers) and result_of (which answer each row is of). Write_RTF() writes it to RTF.

Display rules

A p-value is written to three decimals, p < 0.001 below that and p > 0.999 above, never with stars. Every row is labelled exploratory, with its adjustment named when it has one. An estimate is written to four significant digits with its confidence interval. A statistic R did not compute has no p-value, and its note is R's reason.

Titles and footnotes

The settings title, subtitle and footnotes are text with named placeholders, as for Visualize_GroupComparison(): {measure}, {visits}, {value}, {group}, {n}, {filters}, {date} and {version}. They are the table's attributes title, subtitle and footnotes, the last footnote always the table's own.

See also

Visualize_GroupComparison() and Widget_GroupComparison(), the same view as a figure and as a widget.

Other tables: Table_AssociationScatter(), Table_BiomarkerScreen(), Table_CorrelationMatrix(), Table_CrossTab(), Table_StratifiedSurvival(), Write_RTF()

Examples

Table_GroupComparison(
  Synthetic_Results,
  Synthetic_Participants,
  lSettings = list(
    start_value = "IL-6",
    visits = c("Week 4", "Week 8"),
    value_type = "change",
    baseline_visits = "Baseline",
    group_by = "ARM"
  )
)
#>    Visit          Statistic                  Method
#> 1 Week 4 Test of the groups Welch Two Sample t-test
#> 2 Week 8 Test of the groups Welch Two Sample t-test
#>                                                                                                                                       Estimate
#> 1   Mean (Placebo): 0.02473; Mean (Treatment): -1.21; Difference in means (Placebo - Treatment): 1.235, 95% confidence interval 0.844 to 1.626
#> 2 Mean (Placebo): 0.08592; Mean (Treatment): -1.273; Difference in means (Placebo - Treatment): 1.358, 95% confidence interval 0.9882 to 1.729
#>                             Counts   p-value                     Note
#> 1 Placebo n = 95, Treatment n = 91 p < 0.001 Exploratory, unadjusted.
#> 2 Placebo n = 93, Treatment n = 95 p < 0.001 Exploratory, unadjusted.