Gallery • gsm.bio

Each chart is shown twice, from one list of settings. First the static figure, a ggplot that Visualize_*() returns, for a report or a slide. Then the widget, Widget_*(), which draws the same view in the browser and lets a reader move it.

The two share one list of settings, the title and subtitle with them, and each writes its own footnote last: the widget’s says it was drawn by bio.viz, from results computed by R and stored with the page, and the figure’s that it was drawn by gsm.bio.

Both take their statistics from the same Analyze_*() call on the same rows. The number printed under the figure is the one stored in the widget, so the two say the same thing. The figure’s title, subtitle and footnotes are written with the chart’s placeholders ({measure}, {n}, {filters} and the rest), and the figure always writes its own footnote last: the date it was drawn, by gsm.bio, and R’s method and counts behind each statistic it printed. Everything here is the synthetic study, Synthetic_Results, Synthetic_Participants and Synthetic_Outcomes.

Group comparison

IL-6’s change from Baseline by arm, the difference the study plants, with Welch’s test of each visit.

lSettings <- list(
  start_value = "IL-6", visits = c("Week 4", "Week 8"), value_type = "change", baseline_visits = "Baseline",
  group_by = "ARM", groups = lColumns, filters = lColumns
)
lSettings <- c(lSettings, list(title = "{measure}: {value} by {group}", subtitle = "{n} participants, at {visits}"))
Visualize_GroupComparison(Synthetic_Results, Synthetic_Participants, lSettings)

Box plots of IL-6's change from Baseline at Week 4 and Week 8, Placebo against Treatment, with Welch's test of each visit and the difference in means in each panel's heading.

Widget_GroupComparison(Synthetic_Results, Synthetic_Participants, lSettings = lSettings)

Association scatter

TNF-alpha against IL-10 at Baseline, the correlation the study plants, with R’s linear fit and its band.

lSettings <- list(
  x = list(measure = "TNF-alpha", visit = "Baseline"), y = list(measure = "IL-10", visit = "Baseline"),
  fit = "linear", filters = lColumns
)
lSettings <- c(lSettings, list(title = "{y} against {x}", subtitle = "{n} participants"))
Visualize_AssociationScatter(Synthetic_Results, Synthetic_Participants, lSettings)

Scatter plot of IL-10 against TNF-alpha at Baseline, one point per participant, with R's linear fit and its confidence band; the coefficient and the line's estimates are printed beneath.

Widget_AssociationScatter(Synthetic_Results, Synthetic_Participants, lSettings = lSettings)

Correlation matrix

Every biomarker against every other at Baseline. The figure writes each coefficient above the diagonal, to two decimals as the grid does.

lSettings <- list(visit = "Baseline", filters = lColumns)
lSettings <- c(lSettings, list(title = "{heading}", subtitle = "{variables} biomarkers, {n} participants"))
Visualize_CorrelationMatrix(Synthetic_Results, Synthetic_Participants, lSettings)

Grid of Pearson coefficients between the twelve biomarkers at Baseline, coloured from negative to positive, with each coefficient written above the diagonal.

Widget_CorrelationMatrix(Synthetic_Results, Synthetic_Participants, lSettings = lSettings)

Biomarker screen

Every biomarker’s change to Week 4, Placebo against Treatment, as a standardised difference with both p-values. The widget opens any row in its group comparison.

lSettings <- list(
  visit = "Week 4", value_type = "change", group_by = "ARM", baseline_visits = "Baseline",
  groups = lColumns, filters = lColumns, group_comparison = list(groups = lColumns)
)
lSettings <- c(lSettings, list(title = "{heading}", subtitle = "{biomarkers} biomarkers, {n} participants"))
Visualize_BiomarkerScreen(Synthetic_Results, Synthetic_Participants, lSettings)

Forest plot of the twelve biomarkers' standardised differences in change to Week 4, Placebo less Treatment, largest first, each with its interval and both p-values.

Widget_BiomarkerScreen(Synthetic_Results, Synthetic_Participants, lSettings = lSettings)

Cross-tabulation

Response by CRP at Baseline cut at its median, with R’s chi-square test.

lSettings <- list(
  row_by = "RESPONSE", col_by = list(measure = "CRP", visit = "Baseline", cut = "median"),
  groups = lColumns, filters = lColumns
)
lSettings <- c(lSettings, list(title = "{rows} by {columns}", subtitle = "{n} participants"))
Visualize_CrossTab(Synthetic_Results, Synthetic_Participants, lSettings)

Stacked bars of response by CRP at Baseline cut at its median, as percentages of each response with the counts written in, and R's chi-square test beneath.

Widget_CrossTab(Synthetic_Results, Synthetic_Participants, lSettings = lSettings)

Stratified survival

Event-free survival by CRP at Baseline cut at its median, the survival effect the study plants. The figure draws R’s own survfit() curves with their confidence bands. The widget draws the same curves and lets a reader drag the cut line.

lSettings <- list(
  endpoint = "EFS", group_by = list(measure = "CRP", visit = "Baseline", cut = "median"),
  groups = lColumns, filters = lColumns
)
lSettings <- c(lSettings, list(title = "{endpoint} by {group}", subtitle = "{n} participants"))
Visualize_StratifiedSurvival(Synthetic_Results, Synthetic_Participants, lSettings, dfOutcomes = Synthetic_Outcomes)

Kaplan-Meier curves of event-free survival for CRP at Baseline at or below and above its median, with their confidence bands and censor marks; the log-rank test, the medians and the hazard ratio are printed beneath.

Widget_StratifiedSurvival(Synthetic_Results, Synthetic_Participants, lSettings = lSettings, dfOutcomes = Synthetic_Outcomes)