---
title: "The toxTables function"
author: "Brie Noble, Blake Langlais"
output:
  rmarkdown::html_vignette:
    toc: yes
    toc_depth: 3
vignette: |
  %\VignetteIndexEntry{The toxTables function}
  %\VignetteEncoding{UTF-8}
  %\VignetteEngine{knitr::rmarkdown}
editor_options: 
  markdown: 
    wrap: 72
---

```{r include=FALSE}

knitr::opts_chunk$set(eval = TRUE, message = FALSE, results = 'asis', comment='')
options(width = 200)

```

# Introduction

The `toxTables()` function produces structured statistical tables
similar to those created for CTCAE data. Tables can show arm-level
sample size and frequency distributions for PRO-CTCAE items available in
the provided data frame. P-values from chi-squared or Fisher's exact
tests can also be calculated. Additionally, risk-differences between
arms can be calculated using provided alpha-level. An R data frame is
returned in table format which can be further formatted through R
packages, spreadsheets, or directly copy and pasted into a document.

```{r, load-data}
library(ProAE)
require(knitr)

require(kableExtra)

data(tox_acute)

```

In the examples below, we will use the provided `ProAE::tox_acute` data
frame. This data was simulated to demonstrate a common symptomatic AE
profile where the drug group experiences acute toxicity followed by
symptom abatement over the course of treatment.

In order to use the `toxTables()` function the data frame needs to be in
long format (multiple rows per patient). Additionally, the cycle
variable needs to be numeric. 

# Example 1 - Default Table

```{r, results='markup'}

acute <- tox_acute

str(acute)
  
```

The `toxTables()` function can be used to create a summary table
that calculate the frequency of patients with nausea scores >= 1 and >= 3 by treatment arm.

The default summary measure used is the baseline adjusted score which is
derived by the following:

If the maximum score post-baseline is more severe than the baseline
score, then the maximum score post-baseline is used as the adjusted
score.

Otherwise, if the maximum score post-baseline is the same or less serve
than the baseline score, zero (0) is used as the adjusted score.

```{r}

table_1 <- toxTables(dsn = acute,
                    id_var="id",
                    cycle_var="Cycle",
                    baseline_val = 1,
                    arm="arm")

```

We can use the `knitr::kable()` function to see a simple RMarkdown version
of the table.

```{r}

knitr::kable(table_1$individual)

```


# Example 2 - Customizing tables

Further customization can happen using `kable()` and `kableExtra()` or the data frame can be exported 
to spreadsheet or copy and pasted into a document


```{r}

knitr::kable(table_1$individual,
             col.names = c('Item/Attribute', 'Drug', 'Placebo', 'Drug', 'Placebo', "p", "Drug", "Placebo", "p"),
             caption = "Table 1. Frequency distributions of patients with nausea score >= 1 and >= 3, by treatment arm") %>%               
             kableExtra::add_header_above(c(" ", "n" = 2,  "Score >= 1" = 2, " ", "Score >= 3" = 2, " ")) %>% 
             kableExtra::add_footnote("p: result from a standard chi square test comparing score threshold frequency distributions between arms",
                          notation = "symbol")

```