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

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

```{r setup}
library(measoshapes)
```


```{r patchwork}
library(dplyr)
library(measoshapes)


## using ggplot2 we set up  a basic plot, for re-use
## with subplots over our map - turn off grids, borders
## and background colour
library(ggplot2)
g0 <- ggplot(mtcars) + xlab(NULL) + ylab(NULL) +
  theme_bw() +
  theme( panel.grid.major = element_blank(),
         panel.grid.minor = element_blank(),
         panel.border = element_blank(),
         panel.background = element_blank(),
         rect = element_rect(fill = "#44444411"))

## now use that basis plot and create various data plots
## p1:p5, variously scatterplot, bar chart, boxplot
p1 <- g0 +
  geom_point(aes(mpg, disp))
p2 <- g0 +
  geom_boxplot(aes(gear, disp, group = gear))
p3 <- g0 +
  geom_line(aes(hp, wt, colour = mpg)) +
  theme(legend.position = "none")
p4 <- g0 +
  geom_bar(aes(gear)) +
  facet_wrap(~cyl)
p5 <- g0 +
  geom_bar(aes(cyl))

## create the background map, the regions in their colours
## (half-transparent, so the plots stand out; the temperate regions have no fill)
g <- ggplot(measo_regions06) +
  geom_sf(aes(fill = fill), alpha = 0.5) +
  scale_fill_identity(na.value = "transparent") +
  theme_bw()

## now use patchwork to compose
library(patchwork)

## the layout takes the background map, and then p1:p5
## placing them first in a 10*10 column figure
## and each p1:p5 falls in its own
## t(op), l(eft), r(right), b(ottom) column number or range
layout <- c(
  area(t = 1, l = 1, b = 10, r = 10),

  area(t = 3, l = 5, b =3, r = 5),
  area(t = 4, l = 7, b = 4, r = 7),
  area(t = 6, l = 4, b = 6, r = 4),
  area(t = 8, l = 7, b = 8, r = 7),
  area(t = 8, l = 5, b = 8, r = 5)
)

## finally, draw it
g + p1 + p2 + p3 + p4 + p5 + plot_layout(design = layout)


```