adobe_analytics_audit.Rmd

---
title: "NYL Adobe Analytics Audit"
output:
  html_document:
    toc: yes
    df_print: paged
  pdf_document:
    toc: yes
params:
<<<<<<< HEAD
  company_id: newyora
  rsid: nylgmadlaunchprod
  google_account: chima.umeakunne@gofurther.com
  start_date: 2023-10-01
  end_date: 2023-12-31
  relative_date: FALSE 
  min_instances: 100
=======
  company_id: "[ADD COMPANY ID HERE]"
  rsid: "[ADD REPORT SUITE ID HERE]"
  google_account: "[ADD YOUR GOOGLE ACCOUNT EMAIL ADDRESS HERE]"
>>>>>>> b41c1dae6d9297157b5dff0328e929b0d31f0344
---

```{r setup, include = FALSE, warning = FALSE, message = FALSE}
# Capturing the start time -- the final output will include how long
# it took for the process to run.
monitor_start_time <- Sys.time()

# Interim output will be written to an output directory (for future use; not core to the script),
# so make sure that directory exists.
dir.create("output", showWarnings = FALSE)

# Check for packages needed and then load the packages
if (!require("pacman")) install.packages("pacman")
pacman::p_load(tidyverse,
               adobeanalyticsr,
               googlesheets4,
               scales,
               htmltools,  # Adding logo
               knitr,
               kableExtra,
               anytime)

# Set the email to use when creating the Google Sheet. You can manually run gs4_auth()
# once before running this workbook to get a token generated and then use the command
# below to provide the email address to use. This just keeps the script from needing
# to pause with a prompt when the write_data_to_sheets.R script runs. Simply comment
# out this line if you're cool with being prompted.

googlesheets4::gs4_auth(email = params$google_account)

company_id <- params$company_id
rsid <- params$rsid

# Set the date range for how far back to go. "Last 60 Days" is a good place to start,
# but this can also be coded as a full month or something else.
if(params$relative_date == TRUE) {
  end_date <- Sys.Date() - 1
  start_date <- Sys.Date() - 186 #Change to 60  after CR run. #This year to date
} else {
  start_date <- params$start_date
  end_date <- params$end_date
}

date_range <- c(start_date, end_date) 

# Cutoff for total instances (or occurrences) below which will flag as "Minimal Data"
min_instances <- params$min_instances

# For the summary table, set the thresholds for green and yellow (anything lower than)
# the yellow value--think of it as a %--will be red
summary_good <- 0.9
summary_caution <- 0.7

color_good <- "#00AE4D"
color_caution <- "#FFD400"
color_bad <- "#ED1B23"

# Switch to set to bypass actually pulling data and, instead, simply pull from
# .rds files written out locally on a previous run. Use this if the writing out
# to Google Sheets crapped out on an earlier version. The "time to run" will
# be messed up, but, otherwise, this will work.
use_local_files <- TRUE

# Add logo at top right
htmltools::img(src = knitr::image_uri("sdi_logo_sm.png"), 
               alt = 'logo', 
               style = 'position:absolute; top:0; right:0; padding:10px;')

# Authorize
aw_auth(type = "oauth")

# Get lists of all props, eVars, listvars, events, segments, and calculated metrics
df_dims <- aw_get_dimensions(company_id = company_id, rsid = rsid)
# Address weird case where there may be no "description" field returned. Add a blank one
if("description" %in% colnames(df_dims) == FALSE){
  df_dims$description <- ""
}

# If there are no classifications, then there is no "parent" column, so add one and make
# it FALSE
if("parent" %in% colnames(df_dims) == FALSE){
  df_dims$parent <- NA
}

df_metrics <- aw_get_metrics(company_id = company_id, rsid = rsid)
# Address weird case where there may be no "description" field returned. Add a blank one
if("description" %in% colnames(df_metrics) == FALSE){
  df_metrics$description <- ""
}

```

```{r get_data, message=FALSE, warning=FALSE, include=FALSE}

# Typically, this will run all of the AA queries, but that can take a while, and
# the writing-to-a-Google-Sheet may crap out. In that case, changing the flag
# to TRUE above and running again will simply read the values from locally stored
# versions of the data, which is much faster.
if(use_local_files == TRUE){
  # Get/process data
  source("code/get_evar_data.R")
  source("code/get_events_data.R")
  source("code/get_listvar_data.R")
  source("code/get_prop_data.R")
  source("code/get_traffic_source_data.R")
} else {
  df_evars <- readRDS("output/df_evars.rds")
  df_events <- readRDS("output/df_events.rds")
  df_events_daily <- readRDS( "output/df_events_daily.rds")
  df_listvars <- readRDS("output/df_listvars.rds")
  df_props <- readRDS("output/df_props.rds")
  df_marketing_channels <- readRDS("output/df_marketing_channels.rds")
  df_campaign <- readRDS("output/df_campaign.rds")

  marketing_channel_messages <- readRDS("output/marketing_channel_messages.rds")
  cmp_ttl <- marketing_channel_messages$cmp_ttl
  cmp_class_ttl <- marketing_channel_messages$cmp_class_ttl
  cmp_class_w_data <- marketing_channel_messages$cmp_class_w_data
  mc_low <- marketing_channel_messages$mc_low
  mc_none <- marketing_channel_messages$mc_none
  mc_session_refresh <- marketing_channel_messages$mc_session_refresh
}

```

```{r write_out_the_data, include = FALSE, warning = FALSE, message = FALSE}
# Write the raw data out to Google Sheets
source("code/write_data_to_sheets.R")
```

## Summary

```{r child = "code/01_summary.Rmd"}
```

## eVars

```{r child = "code/02_evars.Rmd"}
```

## listvars

```{r child = "code/03_listvars.Rmd"}
```

## props

```{r child = "code/04_props.Rmd"}
```

## Events

```{r child = "code/05_events.Rmd"}
```

## Traffic Sources

```{r child = "code/06_channels.Rmd"}
```

## Audit Duration

```{r time_to_run, echo = FALSE, warning = FALSE, message = FALSE}

# Check how long it took
monitor_end_time <- Sys.time()
```

This audit took **`r as.numeric(monitor_end_time - monitor_start_time) %>% round(1)` `r (monitor_end_time - monitor_start_time) %>% attributes() %>% .$units`** to complete.