---
title: "Example Usage"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Example Usage}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

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

```{r setup, message=FALSE}
library(airPb)
library(dplyr)
library(tidyr)
library(purrr)
```

### Example 1: Using `calculate_airPb()` and `calculate_scaling_factors()` and manually applying scaling factors to air lead estimates.

```{r}
d <- tibble::tribble(
  ~id,         ~lon,        ~lat,
    809089L, -84.69127387, 39.24710734,
    813233L, -84.47798287, 39.12005904,
    814881L, -84.47123583,  39.2631309,
    799697L, -84.41741798, 39.18541228,
    799698L, -84.41395064, 39.18322447
  )

my_dates <- data.frame(start_date = as.Date(c("2010-01-08", "2012-06-08", "2010-01-09", "2015-04-09", "2010-01-10")),
                       end_date = as.Date(c("2010-02-08", "2012-07-08", "2010-02-09", "2015-05-09", "2010-02-10")))

d %>% 
  mutate(airPb = calculate_airPb(. , return.LU.vars = FALSE), 
         scaling_factors = calculate_scaling_factors(my_dates), 
         scaled_airPb = airPb * scaling_factors)
```

### Example 2: Using the `add_scaled_airPb()` wrapper function to automatically apply scaling factors to air lead estimates.

A common use case is calculating monthly exposures. For example, we may have a pair of coordinates recorded annually for each participant. In the data below, we have 2 unique ids, with lat/lon recorded once per year.

```{r}
d <- tibble::tribble(
  ~id,         ~lon,        ~lat,        ~date,
    809089L, -84.69127387, 39.24710734, as.Date("2010-01-08"),
    809089L, -84.69127387, 39.24710734, as.Date("2011-01-08"),
    809089L, -84.69127387, 39.24710734, as.Date("2012-01-08"),
    799697L, -84.41741798, 39.18541228, as.Date("2011-01-10"),
    799697L, -84.41741798, 39.18541228, as.Date("2012-02-10")
  )
```

We want to scale the air lead measurements to monthly exposures between these dates, but we need `start_date` and `end_date` columns that represent the monthy time periods we want to average over.

```{r}
d <- d %>% 
  mutate(from = date,
         to = from + lubridate::years(1)) %>% 
  group_by(id, date) %>% 
  nest() %>% 
  mutate(dates = map(data, ~seq.Date(from = .x$from, 
                                     to = .x$to, 
                                     by = '3 months'))) %>% 
  unnest(cols=c('data', 'dates')) %>% 
  dplyr::select(-from, -to) %>% 
  rename(start_date = dates) %>% 
  mutate(end_date = lead(start_date)) %>% 
  filter(!is.na(end_date)) %>% 
  ungroup()

d %>% 
  mutate(scaled_airPb = add_scaled_airPb(.))
```