---
title: "Spatial EDA"
output: pdf_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = F)
library(knitr)
library(ggplot2)
library(dplyr)
library(gstat)
library(sp)
library(datasets)
library(SemiPar)
library(ggmap)
set.seed(02192021)
```
# Exploratory Data Analysis

## EDA Overview

\vfill

## Data Decomposition: Time Series

\vfill

```{r}
plot(decompose(AirPassengers))
```

\newpage

## Data Decomposition: Spatial Data

\vfill


## Response Surface vs. Spatial Surface

\vfill

## Spatial EDA Overview

1. Map of locations
2. Histrogram or other distributional figure
3. 3D scatterplot
4. General Regression EDA
5. Variograms and variogram clouds
6. Anistopic diagnostics

\newpage

# Scallops Data Example

## 1. Map of Locations
```{r, message = F}
data(scallop)
scallop <- scallop %>% mutate(log.catch = log(tot.catch + 1))

scallops_box <- c(left = min(scallop$longitude)-.25, bottom = min(scallop$latitude)-.25, right = max(scallop$longitude)+.25, top =max(scallop$latitude)+.25)
scallops_map <- get_map(scallops_box) %>% ggmap()
  
scallops_map + geom_point(aes(x = longitude, y = latitude, color= tot.catch), data = scallop, inherit.aes = F)
```

## 1. Map of Locations - Takeaways

\vfill

\newpage

## 2. Histogram
```{r}
scallop %>% ggplot(aes(x=tot.catch)) + geom_histogram(bins=70) + ggtitle('histogram of scallop catch') + xlab('Number of Scallops') + theme_bw()
```

```{r}
scallop %>% ggplot(aes(x=log.catch)) + geom_histogram(bins=70) + ggtitle('histogram of log scallop catch (+1)') + xlab('Number of Scallops') + theme_bw()
```

## 2. Histogram - Takeaways

\vfill

\newpage

## 3. 3D scatterplot
```{r}
scallop %>%  ggplot(aes(x=longitude, y = latitude)) + geom_point(aes(color=tot.catch)) + scale_colour_gradient2() + theme_bw()
```

```{r}
scallop %>% ggplot(aes(x=longitude, y = latitude)) + geom_point(aes(color=log.catch)) + scale_colour_gradient2() + theme_bw()
```

## 3. 3D scatterplot - Takeaways

\vfill

\newpage

## 4. General Regression EDA

\vfill

## 5. Variograms and variogram clouds

```{r,  eval = T}
coordinates(scallop) = ~longitude+latitude
scallop.sp <- scallop
proj4string(scallop.sp) <- CRS("+proj=longlat +datum=WGS84")  ## for example
scallop.utm <- spTransform(scallop.sp, CRS("+proj=utm +zone=18 ellps=WGS84"))

plot(variogram(log.catch~1, scallop.utm))
```

```{r}
plot(variogram(log.catch~1, scallop.utm, cloud=T))
```

## 5. Variograms and variogram clouds: Takeaways

\vfill

\newpage 

## 6. Anisotropy

_Goal:_ Determine if direction influences spatial structure


## Directional Variogram


```{r}
plot(variogram(log.catch~1, scallop.sp, alpha = c(0, 45, 90, 135)))
```