Group comparison

Does this biomarker differ between these groups? One value across the levels of a category at chosen visits, as boxes, violins or points with the number in each group beneath, a second grouping by colour, panels by one further variable and a logarithmic scale. It opens on a tile for every biomarker, a line per group through the group's median at each scheduled visit, and a click on a tile opens that biomarker across the visits in one picture, with the number in each group and R's test of the groups under each visit; a click on a visit opens it alone. Unscheduled visits are left out until they are switched on. Click a box or a point to list its participants and open a participant's profile. The data is the synthetic study: 200 made-up participants, and no real one.

  • The chart opens on a tile for each biomarker: a line for each group through the group’s median at every visit, in the groups’ colours, with one key above the tiles. Each biomarker has its own value axis, and its range is printed under its tile. An axis is never narrower than 1.25 standard deviations of the results at the baseline visit, so lines that differ by less than that stay close to flat, and a tile whose lines part is one to open. Click a biomarker, or press Enter or Space on it, to view it across the visits; choose All Biomarkers under Biomarker in the controls, or in the trail above the chart, to come back.
  • To find which biomarker differs between two groups, start from the biomarker screen: every biomarker at once, one row each, with R's standardised difference and its p-values adjusted across them. A click on a row there opens that biomarker here.
  • Choose the value, the visits, the groups and the scale in the controls: they apply to every tile, and Tiles draw switches the lines from medians to means. Draw as applies once a biomarker is open, and Colour by and Panel by once a biomarker and a visit are; each says so until then. On a phone the controls are folded away above the chart: tap Controls to open them.
  • With one biomarker open and every visit ticked, the chart draws it over time: the groups side by side at each visit, as boxes, as means with standard errors or as medians with quartiles, and under each visit the number in each group and R’s test of the groups there. Adjust across visits has R adjust those p-values across the visits. Click a visit’s name to view that visit alone; the trail above the chart, or All in the Visit control, leads back.
  • With one visit open, or a few, each visit is a panel. Click a box, a violin or a point to list its participants, and a row of the list to open that participant's profile.
  • With one biomarker open, the row under the visits, and the line under each panel, is a test of the groups, computed by R. The tiles print no test and ask R for nothing. R is started in this browser the first time a biomarker is opened: the row says it is waiting, and the line under it what that first start downloads, until R answers. Nothing leaves this machine, and the chart computes no test itself.
  • Choose the test under Statistics in the controls: a Welch t-test or a Wilcoxon rank-sum test between two groups, a one-way ANOVA or a Kruskal-Wallis test across more, or none. Group by Arm and sex for four groups, and switch on Pairwise comparisons to compare every pair, with the p-values adjusted across the pairs.
  • Every result is exploratory. Each visit's test is its own, adjusted across the visits only when you ask for it, and the row says which; panels are not adjusted for one another. A change to a filter, a group or the test clears the line and asks R again, on the participants then drawn.

The controls, the filters, the listing, the participant profile and the Chart.js this chart draws with are safety.viz's, version 1.9.0, loaded beside bio.viz from a copy with a record of where it came from.

The tests are gsm.bio's, version 0.2.0.9000: R in this browser is given one file of R, copied from gsm.bio at commit 9eda3a8 with a record of where it came from. The same file, run in desktop R on the rows this chart hands over, wrote the answers the browser tests hold this page to.