-
-
Save wolfgang-noichl/10683113 to your computer and use it in GitHub Desktop.
Lenyard
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| # Proof of concept Schleiferlturnier-Teamfinder | |
| # Sehr schlichte Implementierung; Idee: Simulated annealing | |
| # Bisher umgesetzt: lediglich zufaelliges Austauschen bei _besserer_ "Energie", | |
| # d.h. keine Fluktuationen mit hoeherer Energie moeglich. | |
| # Weitere Verbesserungsmoeglichkeit: selektives Update Begegnungsmatrix (nicht | |
| # bei jedem Energieupdate komplett von vorn berechnen!!!) | |
| # Aber es scheint trotzdem einigermassen zu gehen :-D | |
| import time | |
| import numpy as np | |
| import random | |
| from matplotlib import pyplot as pp | |
| pp.ion() | |
| random.seed() | |
| # Potential? Vorerst begegnungen^2 | |
| # Die "Eigenbegegnungen" gehoeren eigentlich raus, und ein sinnvoller Vergleich muesste her! | |
| # Fluktuation? Um Problemen wechselnder Spielanzahl vorerst aus dem | |
| # Weg zu gehen, schlicht Austausch zweier Spieler! | |
| #Anzahl Spieler | |
| N = 40 | |
| #N = 16 | |
| #Anzahl zu spielende Spiele | |
| M = 120 | |
| #M = 30 | |
| #Anzahl Spieler pro Team | |
| L = 4 | |
| #L = 4 | |
| #zufaellige erste Permutationen fuer saemtliche Teams | |
| teams = np.arange(M*L,dtype=int) % N | |
| np.random.shuffle(teams) | |
| teams.shape = (M,L) | |
| begegnungen = np.zeros((N,N), dtype=int) | |
| begegnungen[0][0] = 6 | |
| image = pp.imshow(begegnungen, interpolation="nearest") | |
| pp.show() | |
| pp.colorbar() | |
| for team in teams: | |
| for i in range(L): | |
| for j in range(L): | |
| begegnungen[team[i]][team[j]] += 1 | |
| e_hist = [] | |
| tic = time.clock() | |
| e = np.sum(begegnungen * begegnungen) | |
| for i in range(100000000): | |
| #while True: | |
| if i % 1000 == 0: | |
| begegnungen_plot = np.copy(begegnungen) | |
| begegnungen_plot *= (1-np.identity(N, dtype=int)) | |
| image.set_data(begegnungen_plot) | |
| pp.draw() | |
| team1 = random.randint(0, M-1) | |
| team2 = random.randint(0, M-1) | |
| pos1 = random.randint(0, L-1) | |
| pos2 = random.randint(0, L-1) | |
| if team1 == team2: | |
| continue | |
| if teams[team1][pos1] == teams[team2][pos2]: | |
| continue | |
| delta_e = 0 | |
| for j in range(L): | |
| if j != pos1: | |
| #Faktor 2 wegen verquerem Energiebegriff ("alles doppelt weil Quadrat statt Dreieck") | |
| delta_e -= 2 * (2*begegnungen[teams[team1][pos1]][teams[team1][j]]-1) | |
| delta_e += 2 * (2*begegnungen[teams[team2][pos2]][teams[team1][j]]+1) | |
| if j != pos2: | |
| delta_e -= 2 * (2*begegnungen[teams[team2][pos2]][teams[team2][j]]-1) | |
| delta_e += 2 * (2*begegnungen[teams[team1][pos1]][teams[team2][j]]+1) | |
| if delta_e <= 0: | |
| #austauschen! | |
| p1 = teams[team1][pos1] | |
| p2 = teams[team2][pos2] | |
| teams[team1][pos1] = p2 | |
| teams[team2][pos2] = p1 | |
| e += delta_e | |
| #begegnungen neu berechnen, naja, idealerweise selektiver! | |
| begegnungen = np.zeros((N,N), dtype=int) | |
| for team in teams: | |
| for j in range(L): | |
| for l in range(L): | |
| begegnungen[team[j]][team[l]] += 1 | |
| print i, "," , e | |
| toc = time.clock() | |
| print "Time: ", toc - tic | |
| a = input("Pause") |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment