Top Distance between 1st and 2nd WCA competition

# [email protected]

# program made after a request from Felipe Cardim
# place this program alongside with the tsv (unziped) export from
# https://www.worldcubeassociation.org/results/misc/WCA_export.tsv.zip
# and run it with
# python program.py
# it takes about 24 min on my PC

import csv
from datetime import date
from math import sin, cos, sqrt, atan2, radians

class Location:

    def __init__(self, latitude, longitude):

        self.latitude = latitude
        self.longitude = longitude

    def get_latitude(self):
        return self.latitude

    def get_longitude(self):
        return self.longitude

class Competition:

    def __init__(self, name, date, location):

        self.name = name
        self.date = date
        self.location = location

    def get_name(self):
        return self.name

    def get_date(self):
        return self.date

    def get_location(self):
        return self.location

    def __lt__(self, other):
         return self.date < other.date

    def __str__(self):
        return self.name

def dist(loc1, loc2):
    # https://stackoverflow.com/questions/19412462/getting-distance-between-two-points-based-on-latitude-longitude

    R = 6373.0

    lat1, lon1 = loc1.get_latitude(), loc1.get_longitude()
    lat2, lon2 = loc2.get_latitude(), loc2.get_longitude()

    dlon = lon2 - lon1
    dlat = lat2 - lat1

    a = sin(dlat / 2)**2 + cos(lat1) * cos(lat2) * sin(dlon / 2)**2
    c = 2 * atan2(sqrt(a), sqrt(1 - a))

    return R*c

def dis_comp(competition1, competition2):
    return dist(competition1.get_location(), competition2.get_location())

def new_competition(name):

    with open("WCA_export_Competitions.tsv") as tsvfile:

        tsvreader = csv.reader(tsvfile, delimiter="\t")

        for line in tsvreader:

            if name == line[0]:

                year = int(line[5])
                month = int(line[6])
                day = int(line[7])

                this_date = date(year, month, day)

                latitude = radians(float(line[18])/pow(10,6))
                longitude = radians(float(line[19])/pow(10,6))

                location = Location(latitude, longitude)

                return Competition(name, date, location)

    raise ValueError('Name for competition not found.')

def show_results(labels, results_list, list_of_lists, limit = float("inf"), direction = "asc"):
    """Labels is intuitive. Results_list is the main order list and list_of_lists are other lists that are gonna be ordered as well (eg.: country, name, etc). You can set limit if you wish. Direction = 'asc' if you wanna the lower results first; direction = 'des' if you wanna bigger results first."""

    pos = 1

    prev = None

    out = "Pos"
    for x in labels:
        out += "\t"+x
    print out

    temp = []
    if direction == "asc":
        temp = sorted(zip(results_list, range(len(results_list))))
    elif direction == "des":
        temp = sorted(zip(results_list, range(len(results_list))))[::-1]

    for a, b in temp:
        out = ""
        if a != prev:
            if pos > limit:
                break
            out+="%s)\t"%pos
        else:
            out += "---\t"
        out += str(a)+"\t"
        for x in list_of_lists:
            out += str(x[b])+"\t"
        print out.strip()
        prev = a
        pos += 1

def main():

    with open("WCA_export_Results.tsv") as tsvfile:

        tsvreader = csv.reader(tsvfile, delimiter="\t")

        ids = []
        names = [] # competitors names
        competitions = []

        prev = None
        competition = None

        header = True

        for line in tsvreader:

            if header:
                header = False
                continue

            name = line[0] # competition name
            if name != prev:
                competition = new_competition(name)
            prev = name

            this_id = line[7]

            i = -1
            try:
                i = ids.index(this_id)
            except:
                this_name = line[6]

                ids.append(this_id)
                names.append(this_name)

                competitions.append([competition])

                continue

            flag = True
            for x in competitions[i]: # perhaps I could define the method __eq__ for the class and skip a few lines here
                if competition.get_name() == x.get_name():
                    flag = False
                    break
            if flag:
                competitions[i].append(competition)

        for i in range(len(ids)):
            competitions[i] = sorted(competitions[i])

        dist_list = []
        name_list = []
        comp1_list = []
        comp2_list = []

        for i in range(len(ids)):
            if len(competitions[i]) < 2:
                continue

            comp1 = competitions[i][0]
            comp2 = competitions[i][1]

            dist_list.append(dis_comp(comp1, comp2))
            name_list.append(names[i])
            comp1_list.append(str(comp1))
            comp2_list.append(str(comp2))

        labels = ["Dist (km)", "Name", "Competition 1", "Competition 2"]
        list_of_lists = [name_list, comp1_list, comp2_list]
        limit = 1000
        show_results(labels, dist_list, list_of_lists, limit, "des")

main()