---
title: "Examples for rms Package"
author: "FE Harrell"
date: '`r Sys.Date()`'
output:
  html_document:
    toc: yes
    toc_depth: 3
    number_sections: true
    toc_float: 
      collapsed: false
    code_folding: hide
    theme: cerulean
---
<!--
To copy html output of server: cpaws examples.html R/rms/
-->

# Introduction
## Markdown
This is an R Markdown html document using the template that is [here](http://biostat.mc.vanderbilt.edu/KnitrHtmlTemplate). Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. For more details on using R Markdown see <http://rmarkdown.rstudio.com>.

```{r, results='hide'}
require(rms)
knitrSet(lang='markdown')
```

The following (<code>r hidingTOC(buttonLabel="Outline")</code>) uses the Hmisc `hidingTOC` function to define HTML styles related to a floating table of contents that can be minimized or be collapsed to major outline levels.  For more details see [this](http://biostat.mc.vanderbilt.edu/KnitrHtmlTemplate).

`r hidingTOC(buttonLabel="Outline")`

# Data {.tabset}
## Setup
```{r t3}
getHdata(titanic3)  # Get the dataset from the VU DataSets page
mu <- markupSpecs$html   # markupSpecs is in Hmisc
subtext <- mu$subtext
code    <- mu$code
```


## Data Dictionary
```{r ddict}
html(contents(titanic3), maxlevels=10, levelType='table')
```

## Descriptive Statistics`r subtext('for the', code('titanic3'), 'dataset')`
```{r t3d, height=150}
# Set graphics type so that Hmisc and rms packages use plotly
# Chunk header height=150 is in pixels
# For certain print methods set to use html
options(grType='plotly', prType='html')
s <- summaryM(age + pclass ~ sex, data=titanic3)
html(s)
plot(s)
d <- describe(titanic3)
plot(d)
```

The following doesn't work because it overlays two different legends
```{r sub,height=600,eval=FALSE}
# Try combining two plots into one
p <- plot(d)
plotly::subplot(p[[1]], p[[2]],
                nrows=2, heights=c(.3, .7), which_layout=1)
```

# Logistic Regression Model
```{r lrmt,results='asis'}
dd <- datadist(titanic3); options(datadist='dd')
f <- lrm(survived ~ rcs(sqrt(age),5) * sex, data=titanic3)
print(f)
latex(f)
a <- anova(f)
print(a)
plot(a)
```

```{r summary}
s <- summary(f, age=c(2, 21))
plot(s, log=TRUE)
print(s, dec=2)
```

```{r ggp,fig.height=5,fig.width=6}
ggplot(Predict(f, age, sex), height=500, width=650)  # uses ggplotly()
plotp(Predict(f, age, sex))                          # uses plotly directly
plot(nomogram(f, fun=plogis, funlabel='Prob(survive)'))
```

# Survival Plots for `r mu$code('pbc')` Dataset
Hover over the curves to see particular probability estimates and numbers at risk.  Click on legend components to show/hide components.
```{r pbc,fig.height=6,fig.width=7}
getHdata(pbc)
pbc <- upData(pbc, 
              fu.yrs = fu.days / 365.25,
              units = c(fu.yrs = 'year'))
f <- npsurv(Surv(fu.yrs, status) ~ spiders, data=pbc)
survplotp(f, time.inc=1, times=c(5, 10), fun=function(y) 1 - y)
```

# Computing Environment
`r mu$session()`