augment.Rmd

---
title: "bupaR Docs | Augment logs"
---


```{r echo = F, out.width="25%", fig.align = "right"}
knitr::include_graphics("images/icons/enrich.png")
```

***

# Augment logs

```{r eval = F}
library(bupaverse)
library(dplyr)
```

```{r include = F}
library(bupaverse)
library(dplyr)
```


Enriching an event log with calculated metrics can be done using `augment()`. For example, consider [`trace_length()`](control_flow_analysis.html).

```{r}
traffic_fines %>%
	trace_length(level = "case") 
```
Feeding the resulting table back to `traffic_fines` with `augment()` makes the trace length metric available as a case attribute for further analysis. 

```{r}
traffic_fines %>%
	trace_length(level = "case") %>%
	augment(traffic_fines) %>%
	glimpse()
```

## Adjust names

Using the `prefix` argument, you can add a descriptive prefix to the name of the new variable. In the current example, where the variable is called _absolute_, it might be useful to add the prefix _trace_length_. 

```{r}
traffic_fines %>%
	trace_length(level = "case") %>%
	augment(traffic_fines, prefix = "trace_length") %>%
	glimpse()
```

## Select variables

Some metrics return several variables with information. Say you want to add information on the _processing time_ of each _activity_ to the data. 

```{r}
patients %>% 
	processing_time(level = "activity", units = "hours")
```

Calling `augment` without any further arguments will add all columns, from _min_ until _relative_frequency_ to the data. 

```{r}
patients %>% 
	processing_time(level = "activity", units = "hours") %>%
	augment(patients) %>%
	glimpse()
```

Using the `columns` argument we can specify a selection of columns that we want to use for augmenting the log. For example, say we are only interested in the _mean_ and _median_ processing time. Let's also add a descriptive prefix to these columns.

```{r}
patients %>% 
	processing_time(level = "activity", units = "hours") %>%
	augment(patients, columns = c("mean","median"), prefix = "processing_time") %>%
	glimpse()
```
## Adding multiple metrics

When you want to add multiple metrics, it is imperative to save intermediate updates of the data. Consider the example below. 

```{r}
patients %>%
	trace_length(level = "case") %>%
	augment(patients, prefix = "trace_length") %>%
	trace_coverage(level = "case") %>%
	augment(patients, prefix = "trace_frequency") %>%
	glimpse()
```

As you can see only the [`trace_coverage()`](control_flow_analysis.html) values of the second augment are added, while the first augment is lost. This is because the `patients` data set did not get updated after the first `augment()` call. The proper way would be as follows. 

```{r}
patients %>%
	trace_length(level = "case") %>%
	augment(patients, prefix = "trace_length") -> patients

patients %>%
	trace_coverage(level = "case") %>%
	augment(patients, prefix = "trace_frequency") %>%
	glimpse()
```

```{r include = F}
patients <- eventdataR::patients
```


```{r footer, results = "asis", echo = F}
CURRENT_PAGE <-  stringr::str_replace(knitr::current_input(), ".Rmd",".html")
res <- knitr::knit_expand("_button_footer.Rmd", quiet = TRUE)
res <- knitr::knit_child(text = unlist(res), quiet = TRUE)
cat(res, sep = '\n')
```