---
title: "Figures and tables"
output:
rmarkdown::html_vignette:
toc: true
css: themakr-vignette.css
vignette: >
%\VignetteIndexEntry{Figures and tables}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
resource_files:
- figures/logo.svg
- figures/logo-dark.svg
---
[](https://github.com/CTTIR/zhncommandR/actions/workflows/R-CMD-check.yaml)
[](https://cttir.github.io/zhncommandR/)
[](https://CRAN.R-project.org/package=zhncommandR)
[](https://app.codecov.io/gh/CTTIR/zhncommandR?branch=main)
[](https://cran.r-project.org/package=zhncommandR)
[](https://cran.r-project.org/package=zhncommandR)
[](https://opensource.org/licenses/MIT)
[](https://lifecycle.r-lib.org/articles/stages.html#experimental)
```{r setup, include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>",
fig.width = 7, fig.height = 4.5, dpi = 96
)
library(zhncommandR)
path <- zhn_example_path()
```
Every panel of the `zhncommandR` dashboard is a thin wrapper around an exported
function, so the exact figures and tables you see in the app can be reproduced
in a script or report. This vignette walks through them on the bundled,
100 % synthetic example cohort.
## The tables behind the panels
The three readers turn the tumour-documentation sheets into tidy data frames
(cleaned column names, derived treatment year):
```{r}
cohort <- zhn_read_cohort(path, verbose = FALSE)
cohort |>
dplyr::count(diagnose, name = "Faelle", sort = TRUE) |>
dplyr::slice_head(n = 6) |>
knitr::kable(caption = "Cohort by entity (top 6)")
```
The S3 wrappers print and summarise themselves — a one-line cohort overview:
```{r}
summary(cohort)
```
### OPS-8-544 therapy blocks
```{r}
therapy <- zhn_read_therapy(path, verbose = FALSE)
blocks <- zhn_prepare_therapy_blocks(therapy)
blocks |>
dplyr::count(therapieprotokoll, name = "Bloecke", sort = TRUE) |>
dplyr::slice_head(n = 6) |>
knitr::kable(caption = "Complex-chemotherapy blocks per protocol (top 6)")
summary(blocks)
```
### OPS-1-941 complex diagnostics
```{r}
diag <- zhn_read_diagnostics(path, verbose = FALSE)
dblocks <- zhn_prepare_diagnostic_blocks(diag)
summary(dblocks)
```
### Oncoprint and cytogenetics
The alteration free-text is split and classified into a tidy long table; only
true mutations/variants go to the oncoprint, structural/cytogenetic findings
are tabulated separately.
```{r}
onco <- zhn_parse_oncoprint(cohort)
onco |>
dplyr::count(alteration_class, name = "n", sort = TRUE) |>
knitr::kable(caption = "Oncoprint alterations by class")
cyto <- zhn_parse_cytogenetics(cohort)
cyto |>
dplyr::count(alteration_class, name = "n", sort = TRUE) |>
knitr::kable(caption = "Cytogenetic findings by class")
```
## The figures
All figures share one theme (`zhn_theme()`) in the Hugo Coder look — a single
blue data-ink accent (`zhn_pal$accent`), Inter type and a minimal panel — and
each is a plain `ggplot` object you can extend.
```{r diagnoses}
zhn_plot_diagnoses(cohort, col = "diagnose")
```
```{r cases-by-year}
zhn_plot_cases_by_year(cohort)
```
A publication-ready Kaplan-Meier curve (overall survival) with confidence
interval, censoring marks and a numbers-at-risk table:
```{r km, eval = requireNamespace("ggsurvfit", quietly = TRUE), message = FALSE, warning = FALSE, fig.height = 6}
zhn_plot_km(cohort, time_col = "os", event_col = "death_event",
title = "Gesamtüberleben (synthetisch)")
```
See `vignette("zhncommandR")` for the full set of Kaplan-Meier options
(grouping, log-rank p, Cox hazard ratio with proportional-hazards check,
pairwise comparisons, cumulative-incidence scale).
## Exporting paper-ready files
`zhn_save_plot()` writes any of these figures at 600 dpi (PNG, via `ragg`) or as
a vector PDF (via `cairo_pdf`, with the Inter font embedded). The transparency
toggle from the app maps to the `transparent` argument.
```{r export, eval = FALSE}
p <- zhn_plot_cases_by_year(cohort)
zhn_save_plot(p, "cases_by_year.png", format = "png", width = 8, height = 5)
zhn_save_plot(p, "cases_by_year.pdf", format = "pdf", transparent = TRUE)
```
