03-data.Rmd

# Data

In this section, we provide some ways to obtain and use data for learning purposes. We differentiate between external data (data from an outside source such as a file or web address) and internal data (data that already exists within R).

## External Data

Data may be directly imported into the R environment using packages. Most often, data will take the form of Comma Separated Values (`.csv`), Excel Spreadsheets (`.xlsx`), SAS (`.sas7bdat`) or JSON (`.json`). 

```{r echo=FALSE}
sources <- data.frame(Package = c('readr','readxl','haven','jsonlite'),
                      Formats = c('.csv, .txt', '.xls, .xlsx', '.sas7bdat, .sas7bcat, .xpt','.json'),
                      Reference = c('<a href="https://github.com/rstudio/cheatsheets/blob/main/data-import.pdf">Cheat Sheet</a>',
                  '<a href="https://github.com/rstudio/cheatsheets/blob/main/data-import.pdf">Cheat Sheet</a>',
                  '<a href="https://haven.tidyverse.org/reference/read_sas.html">Package Site</a>',
                  '<a href="https://cran.r-project.org/web/packages/jsonlite/vignettes/json-aaquickstart.html">Vignette</a>'),stringsAsFactors = FALSE)


kableExtra::kable(sources, escape = FALSE)
```

### CDISC Datasets

The <a href="https://advance.phuse.global/display/WEL/Test+Dataset+Factory">PHUSE Test Datasets Factory</a> provides a convenient set CDISC datasets in `xpt` format to experiment with. You may access them through the following github repositories:

+ <a href="https://github.com/phuse-org/TestDataFactory/tree/main/Updated/TDF_ADaM">ADaM Data</a>
+ <a href="https://github.com/phuse-org/TestDataFactory/tree/main/Updated/TDF_SDTM">SDTM Data</a>

Below is an example of using the `haven` package to read in `adsl.xpt` directly from the ADaM repository. Note you can also chose to download these data and read them locally from your computer.

```{r, message = F, warning = F, error = F}
library(haven)

adsl <- read_xpt("https://github.com/phuse-org/TestDataFactory/raw/main/Updated/TDF_ADaM/adsl.xpt")
```

```{r echo = F}
kableExtra::kable(head(adsl, n=5), escape = FALSE, align = 'c')
```

## Internal Data

Internal data sources include those that come bundled with R packages or data you create yourself. 

### Package Data

The `datasets` package comes with your default installation of R and houses many example datasets you can utilize. 

Simply run the following to learn about the different ones included.
```{r, eval = F}
library(help = "datasets")
```

Once you've made a selection, running the following will make that data available in your R session.
```{r, eval = F}
# ToothGrowth looks interesting
data(ToothGrowth)
```

```{r echo = F}
kableExtra::kable(head(ToothGrowth, n=5), escape = FALSE, align = 'c')
```

Using package data is not just limited to the `datasets` package. Many specialized packages will make an effort to include some data to demonstrate how their functions work. For example, the `admiral.test` package has many of the SDTM datasets listed in the above repo.

### Simulating Your Own

In some cases you may wish to simulate your own data. You can do so quite easily using a combination of `sample` and `r*` functions built into R.

Below is an example of simulating a few variables typically found in an ADSL dataset.

```{r}
my_sim_data <- data.frame(subjidn = 1:10,
                          sex = sample(c('M','F'), 10, replace = TRUE),
                          age = round(rnorm(10, mean = 30, sd = 5)),
                          stringsAsFactors = FALSE
)
```

```{r echo = F}
kableExtra::kable(head(my_sim_data, n=5), escape = FALSE, align = 'c')
```