summarizeSCI

summarizeSCI() produces, in one call, the summary table of summaryTable() and one plot per variable: a boxplot for continuous variables and a stacked barplot for categorical and dichotomous ones. The type of each plot follows the type the variable has in the table, so table and plots always agree. The plots use the same rows as the table, and the box of the boxplot spans the quartiles as the table computes them, so it shows the numbers of a "median_IQR" table.

Setup and data

We use the hemoglobin values of the second visit of summarySCI::hemoglobin_data, and add a categorical version of them.

library(summarySCI)
data(hemoglobin_data)

hb2 <- hemoglobin_data[hemoglobin_data$visit_nr == 2, ]
hb2$hb_cat <- cut(hb2$hb, c(-Inf, 120, 140, Inf),
                  labels = c("< 120", "120-140", "> 140"))

Table and plots

By default (type = "both"), the result holds the table and the plots, which are addressed by variable name.

res <- summarizeSCI(hb2, vars = c("hb", "hb_cat"), group = "Arm",
                    labels = list(hb = "Hemoglobin (g/L)",
                                  hb_cat = "Hemoglobin category"))
names(res$plots)
#> [1] "hb"     "hb_cat"
res$table

Characteristic

N1

Arm A
N = 832

N1

Arm B
N = 852

Hemoglobin (g/L)

83

135.0 (102.0, 173.0)

84

133.0 (93.0, 150.0)

Missing

0

1

Hemoglobin category

83

84

< 120

8 (10%)

18 (21%)

120-140

55 (66%)

49 (58%)

> 140

20 (24%)

17 (20%)

Missing

0 (0%)

1 (1%)

1N without missing values

2Median (Min, Max); n (%)

res$plots$hb

res$plots$hb_cat

Use type = "table" or type = "plot" to get only one of them.

plots <- summarizeSCI(hb2, vars = c("hb", "hb_cat"), group = "Arm",
                      type = "plot")$plots

Settings carry over to the plots

The settings of the table are applied to the plots as well:

res_n <- summarizeSCI(hb2, vars = "hb_cat", group = "Arm",
                      stat_cat = "n", missing_percent = FALSE)
res_n$plots$hb_cat

Plot options

Options that only concern the plots go into plot_args. Each plot takes the arguments its plotting function has (see ?boxplotSCI and ?stackedbarplotSCI); an argument that no plot uses gives a warning.

res_opt <- summarizeSCI(hb2, vars = c("hb", "hb_cat"), group = "Arm",
                        type = "plot",
                        plot_args = list(risk_table = TRUE,
                                         group_names_short = c("A", "B")))
res_opt$plots$hb

stat_footnote = TRUE adds a footnote in the bottom-right corner saying what a plot shows, e.g. the box and whiskers of a boxplot or the denominator of the percentages of a stacked bar; a character string gives a text of your own.

res_note <- summarizeSCI(hb2, vars = "hb_cat", group = "Arm", type = "plot",
                         plot_args = list(stat_footnote = TRUE))
res_note$plots$hb_cat

Saving the table and plots

With word_output = TRUE, the table is also saved to a Word document at file_path, exactly as summaryTable() saves it:

summarizeSCI(hb2, vars = c("hb", "hb_cat"), group = "Arm",
             word_output = TRUE, file_path = "hemoglobin_table.docx")

save_plots() writes every plot of a result into its own file. The format ("png", "jpeg", "tiff", "pdf" or "svg"), size in inches and resolution are set with format, width, height and res. Since each plot is redrawn at the requested size, its layout always fits the file.

out <- file.path(tempdir(), "figures")

# one file per variable, named after it
basename(save_plots(res, file_path = out))
#> [1] "hb.png"     "hb_cat.png"

# a single name prefixes the variable names
basename(save_plots(res, file_name = "fig", file_path = out, format = "pdf"))
#> [1] "fig_hb.pdf"     "fig_hb_cat.pdf"

# one name per plot, or a single plot
basename(save_plots(res, file_name = c("hb_box", "hb_bars"), file_path = out))
#> [1] "hb_box.png"  "hb_bars.png"
basename(save_plots(res$plots$hb, file_name = "hemoglobin", file_path = out,
                    width = 6, height = 4))
#> [1] "hemoglobin.png"