03_aba.qmd

---
title: "Area-based metrics"
---

```{r, echo = FALSE, warnings = FALSE}
library(rgl)

r3dDefaults <- rgl::r3dDefaults
m <- structure(c(0.921, -0.146, 0.362, 0, 0.386, 0.482, -0.787, 0, 
                -0.06, 0.864, 0.5, 0, 0, 0, 0, 1), .Dim = c(4L, 4L))
r3dDefaults$FOV <- 50
r3dDefaults$userMatrix <- m
r3dDefaults$zoom <- 0.75

knitr::opts_chunk$set(
  comment =  "#>", 
  collapse = TRUE,
  fig.align = "center")

rgl::setupKnitr(autoprint = TRUE) 
options(lidR.progress = FALSE)
```

## Relevant Resources

[Code](https://github.com/tgoodbody/lidRtutorial/blob/main/code/tutorials/03_aba.R)

-   [lidRbook metrics section](https://r-lidar.github.io/lidRbook/modeling-aba.html)
-   [lidRbook modelling section](https://r-lidar.github.io/lidRbook/modeling-aba.html)

## Overview

This code demonstrates an area-based approach for LiDAR data. Basic usage involves computing mean and max height of points within 10x10 m pixels and visualizing the results. The code shows how to compute multiple metrics simultaneously and use predefined metric sets. Advanced usage introduces user-defined metrics for more specialized calculations.

## Environment

```{r clear warnings, warnings = FALSE, message=FALSE}
# Clear environment
rm(list = ls(globalenv()))

# Load packages
library(lidR)
library(sf)
```

## Basic Usage

We'll cover the basic usage of the `lidR` package to compute metrics from LiDAR data.

```{r load_lidar_data}
# Load LiDAR data, excluding withheld flag points
las <- readLAS(files = "data/MixedEucaNat_normalized.laz", select = "*",  filter = "-set_withheld_flag 0")
```

```{r mean_height_pixel_metrics}
# Compute the mean height of points within 10x10 m pixels
hmean <- pixel_metrics(las = las, func = ~mean(Z), res = 10)
hmean
plot(hmean, col = height.colors(50))
```

```{r max_height_pixel_metrics}
# Compute the max height of points within 10x10 m pixels
hmax <- pixel_metrics(las = las, func = ~max(Z), res = 10)
hmax
plot(hmax, col = height.colors(50))
```

```{r multiple_metrics_pixel_metrics}
# Compute several metrics at once using a list
metrics <- pixel_metrics(las = las, func = ~list(hmax = max(Z), hmean = mean(Z)), res = 10)
metrics
plot(metrics, col = height.colors(50))
```

```{r predefined_metrics_pixel_metrics}
# Simplify computing metrics with predefined sets of metrics
metrics <- pixel_metrics(las = las, func = .stdmetrics_z, res = 10)
metrics
plot(metrics, col = height.colors(50))

# Plot a specific metric from the predefined set
plot(metrics, "zsd", col = height.colors(50))
```

## Advanced Usage with User-Defined Metrics

::: callout-tip
## 3rd party metric packages

`lidR` provides flexibility for users to define custom metrics. Check out 3rd party packages like [`lidRmetrics`](https://github.com/ptompalski/lidRmetrics) for suites of metrics.
:::

```{r user_defined_metric_pixel_metrics}
# Generate a user-defined function to compute weighted mean
f <- function(x, weight) { sum(x*weight)/sum(weight) }

# Compute grid metrics for the user-defined function
X <- pixel_metrics(las = las, func = ~f(Z, Intensity), res = 10)

# Visualize the output
plot(X, col = height.colors(50))
```

## Exercises and Questions

Using:

``` r
las <- readLAS("data/MixedEucaNat_normalized.laz", select = "*",  filter = "-set_withheld_flag 0")
```

#### E1.

Assuming that biomass is estimated using the equation `B = 0.5 * mean Z + 0.9 * 90th percentile of Z` applied on first returns only, map the biomass.

#### E2.

Map the density of ground returns at a 5 m resolution with `pixel_metrics(filter = ~Classification == LASGROUND)`.

#### E3.

Map pixels that are flat (planar) using `stdshapemetrics`. These could indicate potential roads.

## Conclusion

In this tutorial, we covered basic usage of the `lidR` package for computing mean and max heights within grid cells and using predefined sets of metrics. Additionally, we explored the advanced usage with the ability to define user-specific metrics for grid computation.