graph_width = Math.min(600, width - 100);
akl_mean = {
return mean(data, x => x.density, x => x.land_area);
}
akl_mean_100 = {
return mean(data, x => x.density, x => x.density >= 100 && x.land_area);
}
akl_weigted_mean = {
return mean(data, x => x.density, x => x.population);
}
population = {
var p = 0;
data.forEach(x => p += +x.population);
return p;
}
total_area = {
var p = 0;
data.forEach(x => p += x.land_area);
return p;
}
census_aotea_sq_distance_bins = make_bins(
data,
d => d.distance,
density_type == 'pop' ? d => d.population : d => d.land_area,
d => d.density,
{binCount:25, xMax:25});
// approximate position where the town hall is (centroid of SA1 7005303)
town_hall = [1757165.3473740562, 5920139.208033819];
data = {
var map = {};
var path = d3.geoPath();
census_collection.features.forEach((x) =>
{
// in km
var ID = x.properties.ID;
var c = path.centroid(x.geometry);
var d = 0.001 * Math.hypot(town_hall[0] - c[0], town_hall[1] - c[1]);
map[ID] = d;
});
var d = [];
census_input_data.pop_data.forEach(x => {
// stupid hack because I forgot to include land_area in the data.
// Doesn’t work for units with no population
var area = +x.population / +x.density || 0;
d.push({ distance: map[x.ID], land_area: area, ...x});
});
return d;
}
/// weighted mean function. Return 0 if there are no valid values.
function mean(list, accessor, weightAccessor)
{
var aggr = new meanAggregator(accessor, weightAccessor);
list.forEach(e => aggr.add(e));
return aggr.value();
}
meanAggregator = {
function MeanAggregator(accessor, weightAccessor)
{
this.accessor = accessor || (e => e);
this.weightAccessor = weightAccessor || (e => 1);
this.w = 0;
this.x = 0;
}
MeanAggregator.prototype.value = function() { return this.w ? this.x / this.w : 0; }
MeanAggregator.prototype.add = function(e) {
var ew = +this.weightAccessor(e);
var val = +this.accessor(e);
if (ew == ew && val == val)
{
this.x += val * ew;
this.w += ew;
}
};
return MeanAggregator;
}
/// Make bins, aggregating data while taking into account weights
/// for now we always get the weighted mean of the values per bin
function make_bins(data, xAccessor, weightAccessor, valueAccessor, options)
{
options = {binCount: 10, xMax: +Infinity, ...options};
var xMin = Infinity;
var xMax = -Infinity;
data.forEach(e => {
var x = +xAccessor(e);
if (x == x) {
xMin = Math.min(x, xMin);
xMax = Math.max(x, xMax);
}
});
xMax = Math.min(options.xMax, xMax);
// choose bins
var xStep = (xMax - xMin) / options.binCount;
xStep = Math.pow(10, Math.ceil(Math.log10(xStep)));
if (xStep * options.binCount * 0.3 > xMax - xMin) xStep *= .2;
if (xStep * options.binCount * 0.75 > xMax - xMin) xStep *= .5;
var xInvStep = 1 / xStep;
var xBinMinIx = Math.floor(xMin * xInvStep);
var xBinMaxIx = Math.ceil(xMax * xInvStep);
var bins = Array(xBinMaxIx - xBinMinIx).fill().map((_, i) => ({
x: (i + xBinMinIx) * xStep,
n: 0,
_aggr: new meanAggregator(valueAccessor, weightAccessor)
}));
data.forEach(e => {
var x = xAccessor(e);
var x_bin = Math.floor((x - xMin) * xInvStep);
if (x < options.xMax)
{
var bin = bins[x_bin];
bin.n += 1;
bin._aggr.add(e);
}
});
bins.forEach(bin => {
bin.value = bin._aggr.value();
bin.weight = bin._aggr.w;
});
return bins;
}
import { census_input_data, census_collection, param }
with {sa_type as sa_type}
from "@roelandschoukens/mode-share-map"
d3 = require("d3")
VegaLite = require('@observablehq/vega-lite')
Purpose-built for displays of data
Observable is your go-to platform for exploring data and creating expressive data visualizations. Use reactive JavaScript notebooks for prototyping and a collaborative canvas for visual data exploration and dashboard creation.
