def evolve_on_cloud(num_generations, current_pop):
for i in np.arange(num_generations):
print("Generation : "+str(i))
line_up = list(itertools.combinations(current_pop, 2))
#Post each chars deets for each match
for match in np.arange(len(line_up)):
genn = i
matchnu=match
p1n = line_up[match][0]
p1moves = parse_file.read_cmd(p1n)
p1str = parse_file.to_str(p1moves)
p2n = line_up[match][1]
p2moves = parse_file.read_cmd(p2n)
p2str = parse_file.to_str(p2moves)
tags = "Evolve"
data = urllib.parse.urlencode({"matchnum":matchnu,"gen":genn,"p1name":p1n , "p1moves":p1str, "p2name": p2n, "p2moves":p2str, "tag":tags})
data = data.encode('utf-8')
request = urllib.request.Request("http://104.197.115.172")
# adding charset parameter to the Content-Type header.
request.add_header("Content-Type","application/x-www-form-urlencoded;charset=utf-8")
f = urllib.request.urlopen(request, data)
print(f.read().decode('utf-8'))
#after the matches are over we need to pull the reults for that generation
#from the db
#And just to make sure that all the simulations have completed well sleep for some time
print("Wait for 45 seconds.")
time.sleep(45)
print("Updating char sheet...")
#update the char sheet
update_gen_from_db(i)
#Now just update relevent dfs and create new pop
#Sort our current pupulation interms of ELO
df_sorted = list_df_gen[i].sort_values('ELO', ascending = False)
#print(df_sorted.head())
#Get last elo
v = df_sorted.iloc[-1].ELO
ratio.append(v)
#Only replace population if were not on the last generation
if i < num_generations-1:
#hah! Get it?
new_pop = nu_gen(df_sorted, len(df_sorted),i+1)
#replace current pop with new pop
current_pop = new_pop.Name.tolist()
#keep track
list_df_gen.append(new_pop)
#On last generation... do some stuff with the list of dfs
elif i == num_generations-1:
print("DONE!")
plt.scatter(range(0,len(ratio)), ratio)
plt.title("Lowest ELO per Generation")
plt.xlabel("Generation")
plt.ylabel("ELO")
plt.show()
def cloud_cast(line, cur_gen, tag):
match =0
#Post each chars deets for each match
#after assess names we get back a 3d list [[["p1","easy"],[p1,med],..]]
for char in np.arange(len(line)):
#them for each of their test martches
for char_sim in np.arange(len(line[char])):
#then we post the details
genn = cur_gen
matchnu=str(match)
p1n = line[char][char_sim][0]
print(p1n)
p1moves = parse_file.read_cmd(p1n)
p1str = parse_file.to_str(p1moves)
p2n = line[char][char_sim][1]
#dont need to get moves for p2
p2str = p2n+ "moves" ##parse_file.to_str(p2moves)
tags = tag
data = urllib.parse.urlencode({"matchnum":matchnu,"gen":genn,"p1name":p1n , "p1moves":p1str, "p2name": p2n, "p2moves":p2str, "tag":tags})
data = data.encode('utf-8')
request = urllib.request.Request("http://104.197.115.172")
# adding charset parameter to the Content-Type header.
request.add_header("Content-Type","application/x-www-form-urlencoded;charset=utf-8")
try:
print("requesting simulation(exp)")
f = urllib.request.urlopen(request, data)
print(f.read().decode('utf-8'))
except Exception as e:
# Return code error (e.g. 404, 501, ...)
# ...
print(str(e))
match+=1
def experiment_on_cloud(num_generations, current_pop):
for i in np.arange(num_generations):
#Go about normal shiaattt
#first evolve
print("Generation : "+str(i))
line_up = list(itertools.combinations(current_pop, 2))
for match in np.arange(len(line_up)):
genn = i
matchnu=match
p1n = line_up[match][0]
p1moves = parse_file.read_cmd(p1n)
p1str = parse_file.to_str(p1moves)
p2n = line_up[match][1]
p2moves = parse_file.read_cmd(p2n)
p2str = parse_file.to_str(p2moves)
tags = "Evolve"
data = urllib.parse.urlencode({"matchnum":matchnu,"gen":genn,"p1name":p1n , "p1moves":p1str, "p2name": p2n, "p2moves":p2str, "tag":tags})
data = data.encode('utf-8')
request = urllib.request.Request("http://104.197.115.172")
# adding charset parameter to the Content-Type header.
request.add_header("Content-Type","application/x-www-form-urlencoded;charset=utf-8")
#f = urllib.request.urlopen(request, data)
#print(f.read().decode('utf-8'))
try:
print("requesting simulation")
f = urllib.request.urlopen(request, data)
print(f.read().decode('utf-8'))
except Exception as e:
# Return code error (e.g. 404, 501, ...)
# ...
print(str(e))
#after the matches are over we need to pull the reults for that generation
#from the db
#And just to make sure that all the simulations have completed well sleep for some time
print("Wait for 45 seconds.")
time.sleep(45)
print("Updating char sheet...")
#update the char sheet
expected_sims_rr = int(sum(range(1,len(current_pop))))
update_gen_from_db(i,expected_sims_rr)
#Now just update relevent dfs and create new pop
#Sort our current pupulation interms of ELO
df_sorted = list_df_gen[i].sort_values('ELO', ascending = False)
#Get last elo
v = df_sorted.iloc[-1].ELO
ratio.append(v)
#Only replace population if were not on the last generation
if i < num_generations-1:
#hah! Get it?
new_pop,new_kids = nu_gen(df_sorted, len(df_sorted),i+1)
print(new_kids)
#replace current pop with new pop
current_pop = new_pop.Name.tolist()
#NOW assess the generations children/
a_pop = assess_names(new_kids, i+1)
#simulate on cloud
cloud_cast(a_pop, i+1, "Experiment")
#update the charstats
print("Wait for 45 seconds.")
time.sleep(45)
print("Updating STATS sheet...")
#each child vs easy med hard
expected_pause = int(len(new_kids)*3)
gen_stats = pull_stats_from_db(i+1, expected_pause)
list_df_stats.append(gen_stats)
#keep track
list_df_gen.append(new_pop)
#On last generation... do some stuff with the list of dfs
elif i == num_generations-1:
print("DONE!")
plt.scatter(range(0,len(ratio)), ratio)
plt.title("Lowest ELO per Generation")
plt.xlabel("Generation")
plt.ylabel("ELO")
plt.show()