---
title: "Tidy Active User Workflows"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Tidy Active User Workflows}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r setup, include = FALSE, purl=FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  eval = FALSE,
  purl = FALSE
)
```

```{r, purl=FALSE}
library(sensortowerR)
library(dplyr)
library(tidyr)
```

## Overview

`st_active_users()` is a focused, tidy wrapper around `st_batch_metrics()` that returns DAU/WAU/MAU in long format.

This makes downstream analysis intuitive with standard tidyverse verbs.

## Basic Pipeline

```{r basic, purl=FALSE}
active_users <- st_active_users(
  os = "unified",
  app_list = c("5ba4585f539ce75b97db6bcb", "67ec0bf3e540b65904256cc4"),
  metrics = c("dau", "mau"),
  date_range = list(start_date = "2025-01-01", end_date = "2025-12-31"),
  countries = c("US", "GB"),
  granularity = "monthly"
)

active_summary <- active_users %>%
  group_by(original_id, app_name, metric, date, country) %>%
  summarise(total_active_users = sum(value, na.rm = TRUE), .groups = "drop")
```

## Wide Reporting Table

```{r wide, purl=FALSE}
active_wide <- active_users %>%
  group_by(original_id, app_name, date, country, metric) %>%
  summarise(value = sum(value, na.rm = TRUE), .groups = "drop") %>%
  pivot_wider(names_from = metric, values_from = value)
```

## Join With Sales Data

```{r join-sales, purl=FALSE}
sales <- st_batch_metrics(
  os = "unified",
  app_list = c("5ba4585f539ce75b97db6bcb", "67ec0bf3e540b65904256cc4"),
  metrics = c("revenue", "downloads"),
  date_range = list(start_date = "2025-01-01", end_date = "2025-12-31"),
  countries = c("US", "GB"),
  granularity = "monthly"
)

sales_wide <- sales %>%
  group_by(original_id, app_name, date, country, metric) %>%
  summarise(value = sum(value, na.rm = TRUE), .groups = "drop") %>%
  pivot_wider(names_from = metric, values_from = value)

combined <- active_wide %>%
  left_join(sales_wide, by = c("original_id", "app_name", "date", "country"))
```

## App-Level KPI Aggregation

```{r kpi, purl=FALSE}
app_kpis <- combined %>%
  mutate(
    rev_per_dau = if_else(!is.na(dau) & dau > 0, revenue / dau, NA_real_),
    rev_per_mau = if_else(!is.na(mau) & mau > 0, revenue / mau, NA_real_)
  ) %>%
  group_by(original_id, app_name, country) %>%
  summarise(
    avg_dau = mean(dau, na.rm = TRUE),
    avg_mau = mean(mau, na.rm = TRUE),
    total_revenue = sum(revenue, na.rm = TRUE),
    avg_rev_per_dau = mean(rev_per_dau, na.rm = TRUE),
    avg_rev_per_mau = mean(rev_per_mau, na.rm = TRUE),
    .groups = "drop"
  )
```