mirror of
https://github.com/titanscouting/tra-superscript.git
synced 2024-11-09 22:44:44 +00:00
309 lines
8.8 KiB
Python
309 lines
8.8 KiB
Python
import abc
|
|
import data as d
|
|
import signal
|
|
import numpy as np
|
|
from tra_analysis import Analysis as an
|
|
from tqdm import tqdm
|
|
|
|
class Module(metaclass = abc.ABCMeta):
|
|
|
|
@classmethod
|
|
def __subclasshook__(cls, subclass):
|
|
return (hasattr(subclass, '__init__') and
|
|
callable(subclass.__init__) and
|
|
hasattr(subclass, 'validate_config') and
|
|
callable(subclass.validate_config) and
|
|
hasattr(subclass, 'run') and
|
|
callable(subclass.run)
|
|
)
|
|
@abc.abstractmethod
|
|
def __init__(self, config, apikey, tbakey, timestamp, competition, *args, **kwargs):
|
|
raise NotImplementedError
|
|
@abc.abstractmethod
|
|
def validate_config(self, *args, **kwargs):
|
|
raise NotImplementedError
|
|
@abc.abstractmethod
|
|
def run(self, *args, **kwargs):
|
|
raise NotImplementedError
|
|
|
|
class Match (Module):
|
|
|
|
config = None
|
|
apikey = None
|
|
tbakey = None
|
|
timestamp = None
|
|
competition = None
|
|
|
|
data = None
|
|
results = None
|
|
|
|
def __init__(self, config, apikey, tbakey, timestamp, competition):
|
|
self.config = config
|
|
self.apikey = apikey
|
|
self.tbakey = tbakey
|
|
self.timestamp = timestamp
|
|
self.competition = competition
|
|
|
|
def validate_config(self):
|
|
return True, ""
|
|
|
|
def run(self):
|
|
self._load_data()
|
|
self._process_data()
|
|
self._push_results()
|
|
|
|
def _load_data(self):
|
|
self.data = d.load_match(self.apikey, self.competition)
|
|
|
|
def _simplestats(self, data_test):
|
|
|
|
signal.signal(signal.SIGINT, signal.SIG_IGN)
|
|
|
|
data = np.array(data_test[3])
|
|
data = data[np.isfinite(data)]
|
|
ranges = list(range(len(data)))
|
|
|
|
test = data_test[2]
|
|
|
|
if test == "basic_stats":
|
|
return an.basic_stats(data)
|
|
|
|
if test == "historical_analysis":
|
|
return an.histo_analysis([ranges, data])
|
|
|
|
if test == "regression_linear":
|
|
return an.regression(ranges, data, ['lin'])
|
|
|
|
if test == "regression_logarithmic":
|
|
return an.regression(ranges, data, ['log'])
|
|
|
|
if test == "regression_exponential":
|
|
return an.regression(ranges, data, ['exp'])
|
|
|
|
if test == "regression_polynomial":
|
|
return an.regression(ranges, data, ['ply'])
|
|
|
|
if test == "regression_sigmoidal":
|
|
return an.regression(ranges, data, ['sig'])
|
|
|
|
def _process_data(self):
|
|
|
|
tests = self.config["tests"]
|
|
data = self.data
|
|
|
|
input_vector = []
|
|
|
|
for team in data:
|
|
|
|
for variable in data[team]:
|
|
|
|
if variable in tests:
|
|
|
|
for test in tests[variable]:
|
|
|
|
input_vector.append((team, variable, test, data[team][variable]))
|
|
|
|
self.data = input_vector
|
|
self.results = []
|
|
for test_var_data in self.data:
|
|
self.results.append(self._simplestats(test_var_data))
|
|
|
|
def _push_results(self):
|
|
|
|
short_mapping = {"regression_linear": "lin", "regression_logarithmic": "log", "regression_exponential": "exp", "regression_polynomial": "ply", "regression_sigmoidal": "sig"}
|
|
|
|
class AutoVivification(dict):
|
|
def __getitem__(self, item):
|
|
try:
|
|
return dict.__getitem__(self, item)
|
|
except KeyError:
|
|
value = self[item] = type(self)()
|
|
return value
|
|
|
|
result_filtered = self.results
|
|
input_vector = self.data
|
|
|
|
return_vector = AutoVivification()
|
|
|
|
i = 0
|
|
|
|
for result in result_filtered:
|
|
|
|
filtered = input_vector[i][2]
|
|
|
|
try:
|
|
short = short_mapping[filtered]
|
|
return_vector[input_vector[i][0]][input_vector[i][1]][input_vector[i][2]] = result[short]
|
|
except KeyError: # not in mapping
|
|
return_vector[input_vector[i][0]][input_vector[i][1]][input_vector[i][2]] = result
|
|
|
|
i += 1
|
|
|
|
self.results = return_vector
|
|
|
|
d.push_match(self.apikey, self.competition, self.results)
|
|
|
|
class Metric (Module):
|
|
|
|
config = None
|
|
apikey = None
|
|
tbakey = None
|
|
timestamp = None
|
|
competition = None
|
|
|
|
data = None
|
|
results = None
|
|
|
|
def __init__(self, config, apikey, tbakey, timestamp, competition):
|
|
self.config = config
|
|
self.apikey = apikey
|
|
self.tbakey = tbakey
|
|
self.timestamp = timestamp
|
|
self.competition = competition
|
|
|
|
def validate_config(self):
|
|
return True, ""
|
|
|
|
def run(self):
|
|
self._load_data()
|
|
self._process_data()
|
|
self._push_results()
|
|
|
|
def _load_data(self):
|
|
self.last_match = d.get_analysis_flags(self.apikey, 'metrics_last_match')['metrics_last_match']
|
|
print("Previous last match", self.last_match)
|
|
self.data = d.pull_new_tba_matches(self.tbakey, self.competition, self.last_match)
|
|
|
|
def _process_data(self):
|
|
|
|
self.results = {}
|
|
self.match = self.last_match
|
|
matches = self.data
|
|
red = {}
|
|
blu = {}
|
|
for match in tqdm(matches, desc="Metrics"): # grab matches and loop through each one
|
|
self.match = max(self.match, int(match['match']))
|
|
red = d.load_metric(self.apikey, self.competition, match, "red", self.config["tests"]) # get the current ratings for red
|
|
blu = d.load_metric(self.apikey, self.competition, match, "blue", self.config["tests"]) # get the current ratings for blue
|
|
|
|
gl2_red_score_total = 0
|
|
gl2_blu_score_total = 0
|
|
|
|
gl2_red_rd_total = 0
|
|
gl2_blu_rd_total = 0
|
|
|
|
gl2_red_vol_total = 0
|
|
gl2_blu_vol_total = 0
|
|
|
|
for team in red: # for each team in red, add up gl2 score components
|
|
|
|
gl2_red_score_total += red[team]["gl2"]["score"]
|
|
gl2_red_rd_total += red[team]["gl2"]["rd"]
|
|
gl2_red_vol_total += red[team]["gl2"]["vol"]
|
|
|
|
for team in blu: # for each team in blue, add up gl2 score components
|
|
|
|
gl2_blu_score_total += blu[team]["gl2"]["score"]
|
|
gl2_blu_rd_total += blu[team]["gl2"]["rd"]
|
|
gl2_blu_vol_total += blu[team]["gl2"]["vol"]
|
|
|
|
|
|
red_gl2 = {"score": gl2_red_score_total / len(red), "rd": gl2_red_rd_total / len(red), "vol": gl2_red_vol_total / len(red)} # average the scores by dividing by 3
|
|
blu_gl2 = {"score": gl2_blu_score_total / len(blu), "rd": gl2_blu_rd_total / len(blu), "vol": gl2_blu_vol_total / len(blu)} # average the scores by dividing by 3
|
|
|
|
|
|
if match["winner"] == "red": # if red won, set observations to {"red": 1, "blu": 0}
|
|
|
|
observations = {"red": 1, "blu": 0}
|
|
|
|
elif match["winner"] == "blue": # if blue won, set observations to {"red": 0, "blu": 1}
|
|
|
|
observations = {"red": 0, "blu": 1}
|
|
|
|
else: # otherwise it was a tie and observations is {"red": 0.5, "blu": 0.5}
|
|
|
|
observations = {"red": 0.5, "blu": 0.5}
|
|
|
|
|
|
new_red_gl2_score, new_red_gl2_rd, new_red_gl2_vol = an.Metric().glicko2(red_gl2["score"], red_gl2["rd"], red_gl2["vol"], [blu_gl2["score"]], [blu_gl2["rd"]], [observations["red"], observations["blu"]]) # calculate new scores for gl2 for red
|
|
new_blu_gl2_score, new_blu_gl2_rd, new_blu_gl2_vol = an.Metric().glicko2(blu_gl2["score"], blu_gl2["rd"], blu_gl2["vol"], [red_gl2["score"]], [red_gl2["rd"]], [observations["blu"], observations["red"]]) # calculate new scores for gl2 for blue
|
|
|
|
red_gl2_delta = {"score": new_red_gl2_score - red_gl2["score"], "rd": new_red_gl2_rd - red_gl2["rd"], "vol": new_red_gl2_vol - red_gl2["vol"]} # calculate gl2 deltas for red
|
|
blu_gl2_delta = {"score": new_blu_gl2_score - blu_gl2["score"], "rd": new_blu_gl2_rd - blu_gl2["rd"], "vol": new_blu_gl2_vol - blu_gl2["vol"]} # calculate gl2 deltas for blue
|
|
|
|
for team in red: # for each team on red, add the previous score with the delta to find the new score
|
|
|
|
red[team]["gl2"]["score"] = red[team]["gl2"]["score"] + red_gl2_delta["score"]
|
|
red[team]["gl2"]["rd"] = red[team]["gl2"]["rd"] + red_gl2_delta["rd"]
|
|
red[team]["gl2"]["vol"] = red[team]["gl2"]["vol"] + red_gl2_delta["vol"]
|
|
|
|
for team in blu: # for each team on blue, add the previous score with the delta to find the new score
|
|
|
|
blu[team]["gl2"]["score"] = blu[team]["gl2"]["score"] + blu_gl2_delta["score"]
|
|
blu[team]["gl2"]["rd"] = blu[team]["gl2"]["rd"] + blu_gl2_delta["rd"]
|
|
blu[team]["gl2"]["vol"] = blu[team]["gl2"]["vol"] + blu_gl2_delta["vol"]
|
|
|
|
temp_vector = {}
|
|
temp_vector.update(red) # update the team's score with the temporay vector
|
|
temp_vector.update(blu)
|
|
|
|
self.results[match['match']] = temp_vector
|
|
|
|
d.push_metric(self.apikey, self.competition, temp_vector) # push new scores to db
|
|
print("New last match", self.match)
|
|
d.set_analysis_flags(self.apikey, 'metrics_last_match', {'metrics_last_match': self.match})
|
|
def _push_results(self):
|
|
pass
|
|
|
|
class Pit (Module):
|
|
|
|
config = None
|
|
apikey = None
|
|
tbakey = None
|
|
timestamp = None
|
|
competition = None
|
|
|
|
data = None
|
|
results = None
|
|
|
|
def __init__(self, config, apikey, tbakey, timestamp, competition):
|
|
self.config = config
|
|
self.apikey = apikey
|
|
self.tbakey = tbakey
|
|
self.timestamp = timestamp
|
|
self.competition = competition
|
|
|
|
def validate_config(self):
|
|
return True, ""
|
|
|
|
def run(self):
|
|
self._load_data()
|
|
self._process_data()
|
|
self._push_results()
|
|
|
|
def _load_data(self):
|
|
self.data = d.load_pit(self.apikey, self.competition)
|
|
|
|
def _process_data(self):
|
|
tests = self.config["tests"]
|
|
return_vector = {}
|
|
for team in self.data:
|
|
for variable in self.data[team]:
|
|
if variable in tests:
|
|
if not variable in return_vector:
|
|
return_vector[variable] = []
|
|
return_vector[variable].append(self.data[team][variable])
|
|
|
|
self.results = return_vector
|
|
|
|
def _push_results(self):
|
|
d.push_pit(self.apikey, self.competition, self.results)
|
|
|
|
class Rating (Module):
|
|
pass
|
|
|
|
class Heatmap (Module):
|
|
pass
|
|
|
|
class Sentiment (Module):
|
|
pass |