Writing spatial samples to shapefile

Users sometimes simulate quantities that do not always correspond to continuous spatial values. Instead, these quantities can be discrete and result from a sampling process across the landscape. In such cases, users often want to visually verify the sampling scheme.

One way to achieve this is by exporting the samples as a shapefile and employing various software tools like Quetzal-CRUMBS or the raster package in R to visualize them.

Input

#include "quetzal/geography.hpp"
#include <cassert>
#include <filesystem>
#include <iostream>
#include <random>
 
using namespace quetzal;
 
int main()
{
    using time_type = int;
    using raster_type = geography::raster<time_type>;
    using latlon = typename raster_type::latlon;
 
    // The raster has 10 bands that we will assign to 2001 ... 2011.
    std::vector<time_type> times(10);
    std::iota(times.begin(), times.end(), 2001);
 
    auto input_file = std::filesystem::current_path() / "data/bio1.tif";
    auto output_dir = std::filesystem::current_path() / "my_sample";
 
    // Read the raster
    auto bio1 = raster_type::from_file(input_file, times);
 
    // Sampling points:
    auto p1 = latlon(48., 0.);
    auto p2 = latlon(51, 3.5);
    auto p3 = latlon(38.0, 9.0);
 
    bio1.to_shapefile({p1, p2, p3}, output_dir);
    std::cout << (std::filesystem::is_directory(output_dir) ? "success" : "error") << std::endl;
 
    // Clean up
    std::filesystem::remove_all(output_dir);
}

Output

success