Visualizing Summary Data

Published Data

The data has been made available in a csv data file named remdesivirfig3.csv

library(ggquickeda) #load ggquickeda
remdesivirdata <- read.csv("./remdesivirfig3.csv") # in vignette folder
knitr::kable(remdesivirdata)

Load the Data into the app

# from R launch the app with the data 
#run_ggquickeda(remdesivirdata) 
# if you have access the the app on a server browse to the file and load it

X/Y Mappings and Splitting Options

• After the app finishes loading the data:
  • Change the mapped y variable(s) to ‘Subgroupvalue’
    
  • Change the mapped x variable(s) to ‘RRLCI’, ‘Recovery Rate Ratio’ and ‘RRUCI’

Summary Data Mapping

• Go to Color/Group/Split/Size/Fill Mappings below the graph
  • cancel the automatic Extra Column Split and set it to ‘None’
  • set Select Row Split to ‘Subgroup’.

Graph Splitting

Facets Options

• Go to Graph Options tab and select the Facets Options subtab. We will be using advanced Faceting (graph splitting options):
  • Set Facet Scales to ‘free_y’ (this allows only Subgroupvalue specific to the Subgroup to appear on each graph panel)
  • Set Facet Spaces to ‘free_y’ (this allows for each panel to occupy space proportional to the number of unique Subgroupvalue)
  • Set Facet Switch to Near Axis to ‘y’ (this will move the y strips to the left)
  • Set Strip Placement to ‘outside’ (this will put the y strips to the left of the axis)

We still have to set text formatting options using the group of subtabs in the lower part of the page:

• Default subtab: Facet Text Size/Angle/Face
  • Set Facet Text Angle Y to 0 (if not automatically set)
  • Check the Y strip Text Bold
• Second subtab: Facet Text Justification
  • Set Facet Text Vertical Justification Y to 1
  • Set Facet Text Horizontal Justification Y to 0
    (not shown in the screenshot)
• Fourth subtab: Facet Labels
  • Set Facet Label to ‘Value(s)’
    (not shown in the screenshot)

At this point you should have this graph:

Facet Options

Ordering of Variables and Values

• Go to main Inputs tab and select the Recode/Reorder Categories subtab.
  • Set Select a variable(s) to ‘Subgroup’
  • Drag and drop the order of levels as in the screenshot

reordering of subgroup

While you can add another variable and manually drag and drop we will demo next another possibility to reorder yvalues using a statistic (e.g. median) of another variable (Subroupvalueorder):

• Select the Reorder Facets or axis Levels subtab
  • Set Reorder This Variable to ‘yvalues’
  • Keep By the: to ‘Median’
  • Set Of this Variable: to ‘Subroupvalueorder’

reordering of Subgroupvalue

Remove Default Points and add a Point Interval

• Select the Points, Lines subtab
  • Set Points to ‘None’
• Select the Median (PIs) subtab
  • Set Median to ‘Median/PI’
  • Set PI Geom: to ‘errorbar’
  • Set PI %: to 1 (this ensures that the errorbar goes from the min to max of values)
  • Uncheck Show lines and check Show points
  • Set Median(s) Point(s) Size: to 6

Median/PI options

Setting Titles, Captions and Logging the X axis

• Select the Graph Options subtab next to the Inputs and in the first default subtab X/Y Axes Log/Labels:
  • Type in one space in the Y axis label (this removes the default label)
  • Type in one space in the X axis label (this removes the default label)
  • Type in “Time to Recovery According to Subgroup” in the Plot Title field
  • Type in “The widths of the confidence intervals have not been adjusted for multiplicity\nand therefore cannot be used to infer treatment effects.\nRace and ethnic group were reported by the patients.” in the Plot Caption field (notice how the\n produce a line break)
  • Set X Axis scale to ‘log10’

And now you should get the below plot !:

Example of what is Possible with ggquickeda

As an example of even more advanced features consider the screenshot below where the Intervals Values are shown while the point Size is proportional to the N of patients. Some theme adjustments to customize the plot and legend were also done.

more options