---
title: "Step 4: Obtain aggregated data on temporal symmetry"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{a05_Summarise_temporal_symmetry}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  warning = FALSE,
  message = FALSE,
  collapse = TRUE,
  comment = "#>",
  fig.width = 7,
  fig.height = 5,
  eval = Sys.getenv("$RUNNER_OS") != "macOS"
)
```

```{r, include = FALSE}
if (Sys.getenv("EUNOMIA_DATA_FOLDER") == "") Sys.setenv("EUNOMIA_DATA_FOLDER" = tempdir())
if (!dir.exists(Sys.getenv("EUNOMIA_DATA_FOLDER"))) dir.create(Sys.getenv("EUNOMIA_DATA_FOLDER"))
if (!CDMConnector::eunomiaIsAvailable()) CDMConnector::downloadEunomiaData()
```

# Introduction
In this vignette we will explore the functionality and arguments of `summariseTemporalSymmetry()` function. This function uses `cdm$intersect` introduced in the previous vignette **Step 1. Generate a sequence cohort** to produce aggregated statistics containing the frequency for different time gaps between the initiation of the marker and the initiation of the index (`marker_date` $-$ `index_date`). The work of this function is best illustrated via an example. 

```{r message= FALSE, warning=FALSE, include=FALSE}
# Load libraries
library(CDMConnector)
library(dplyr)
library(DBI)
library(CohortSymmetry)
library(duckdb)
library(DrugUtilisation)

# Connect to the database
db <- DBI::dbConnect(duckdb::duckdb(), 
                     dbdir = CDMConnector::eunomiaDir())
cdm <- cdmFromCon(
  con = db,
  cdmSchema = "main",
  writeSchema = "main"
)

# Generate cohorts
cdm <- DrugUtilisation::generateIngredientCohortSet(
  cdm = cdm,
  name = "aspirin",
  ingredient = "aspirin")

cdm <- DrugUtilisation::generateIngredientCohortSet(
  cdm = cdm,
  name = "acetaminophen",
  ingredient = "acetaminophen")
```

Recall that in the previous vignette, we've used `cdm$aspirin` and `cdm$acetaminophen` to generate `cdm$intersect` like so:

```{r message= FALSE, warning=FALSE}
# Generate a sequence cohort
cdm <- generateSequenceCohortSet(
  cdm = cdm,
  indexTable = "aspirin",
  markerTable = "acetaminophen",
  name = "intersect",
  combinationWindow = c(0,Inf))
```

# Obtaining temporal symmetry

```{r message= FALSE, warning=FALSE}
summariseTemporalSymmetry(cohort = cdm$intersect) |> 
  dplyr::glimpse()
```

The default unit of the difference of two initiations is measured in months. In this example, the first row is showing there are $6$ cases of *index* happening **after** *marker* with the gap being $29$ months whereas the second row is showing there are $7$ cases of *index* happening **before** *marker* with the gap being $40$ months.

## Modify the cohort based on `cohort_definition_id`

This parameter is used to subset the cohort table inputted to the `summariseTemporalSymmetry()`. Imagine the user only wants to include `cohort_definition_id` $= 1$ from `cdm$intersect` in the `summariseTemporalSymmetry()`, then one could do the following:

```{r message= FALSE, warning=FALSE}
summariseTemporalSymmetry(cohort = cdm$intersect,
                          cohortId = 1) |> 
  dplyr::glimpse()
```

Of course and once again this does nothing because every entry in `cdm$intersect` has `cohort_definition_id` $= 1$.

## Modify `timescale`

Recall the default for the timescale is `month`, one could also change this to either `day` or `year`.

```{r message= FALSE, warning=FALSE}
summariseTemporalSymmetry(cohort = cdm$intersect,
                          timescale = "day") |> 
  dplyr::glimpse()
```

```{r message= FALSE, warning=FALSE}
summariseTemporalSymmetry(cohort = cdm$intersect,
                          timescale = "year") |> 
  dplyr::glimpse()
```

```{r message= FALSE, warning=FALSE, eval=FALSE}
CDMConnector::cdmDisconnect(cdm = cdm)
```