Simple Decision Tree

decision.go

package main

import (
	"bytes"
	"encoding/csv"
	// "encoding/json"
	"fmt"
	"io/ioutil"
	"strings"
)

func main() {
	data, _ := ioutil.ReadFile("raw.csv")

	r := csv.NewReader(bytes.NewReader(data))
	result, _ := r.ReadAll()

	spamScan1(result)
	spamScan2(result)

	// j, _ := json.MarshalIndent(h, "", "\t")
	// fmt.Println(string(j))
}

func spamScan1(dataset interface{}) {
	switch t := dataset.(type) {
	case [][]string:
		h := hashify(t)
		spamScan1(h)
	case []map[string]string:
		fmt.Println("\nspamScan1 -------------")
		for _, row := range t {
			if row["contains_images"] == "true" {
				if row["suspicious_words"] == "true" {
					fmt.Printf("\t%s is SPAM\n", row["id"])
				} else {
					fmt.Printf("\t%s is HAM\n", row["id"])
				}
			} else {
				if row["unknown_sender"] == "true" {
					fmt.Printf("\t%s is SPAM\n", row["id"])
				} else {
					fmt.Printf("\t%s is HAM\n", row["id"])
				}
			}
		}
	}
}

func spamScan2(dataset interface{}) {
	switch t := dataset.(type) {
	case [][]string:
		h := hashify(t)
		spamScan2(h)
	case []map[string]string:
		fmt.Println("\nspamScan2 -------------")
		for _, row := range t {
			if row["suspicious_words"] == "true" {
				fmt.Printf("\t%s is SPAM\n", row["id"])
			} else {
				fmt.Printf("\t%s is HAM\n", row["id"])
			}
		}
	}
}

func hashify(dataset [][]string) []map[string]string {
	headers := []string{}
	for _, d := range dataset[0] {
		headers = append(headers, strings.TrimSpace(d))
	}

	hash := []map[string]string{}

	for _, row := range dataset[1:] {
		m := map[string]string{}
		for n, val := range row {
			m[headers[n]] = strings.TrimSpace(val)
		}
		hash = append(hash, m)
	}

	return hash
}

raw.csv