mid_cardinality_simulation.go

import (
	"log"
	"math/rand"
	"testing"

	"github.com/montanaflynn/stats"
	"gonum.org/v1/gonum/floats"
)

func TestMidCardinalitySimulation(t *testing.T) {
	rnd := rand.New(rand.NewSource(1))
	zipf := rand.NewZipf(rnd, 1.1, 1, 1000)
	var vs []float64
	for k := 0; k < 1000; k++ {
		var sums []float64
		used := make(map[uint64]bool)
		for j := 0; j < 7; j++ {
			var vs []float64
			for len(used) < (j+1)*125 {
				v := zipf.Uint64()
				_, ok := used[v]
				if !ok {
					vs = append(vs, float64(v))
					used[v] = true
				}
			}
			sum, _ := stats.Sum(vs)
			sums = append(sums, sum)
		}
		lastSum := uint64(0)
		for v := uint64(0); v < 1000; v++ {
			_, ok := used[v]
			if !ok {
				lastSum += v
			}
		}
		sums = append(sums, float64(lastSum))
		vs = append(vs, floats.Min(sums)/floats.Max(sums))
	}
	data := stats.Float64Data(vs)
	mean, err := stats.Mean(data)
	if err != nil {
		log.Fatal(err)
	}
	sdev, err := stats.StdDevS(data)
	if err != nil {
		log.Fatal(err)
	}
	log.Printf("mean: %f, stddev: %f, max: %f", mean, sdev, floats.Max(vs))
}