def mlp_classifier_new(config_mlp, X, y, file):
clf = MLPClassifier(hidden_layer_sizes=(config_mlp), max_iter=1000)
skf = StratifiedKFold(n_splits=10)
skf.get_n_splits(X, y)
score = []
f1Score = []
for train_index, test_index in skf.split(X, y):
#print("TRAIN:", train_index, "TEST:", test_index)
score.append(
clf.fit(X[train_index],
y[train_index]).score(X[test_index],
y[test_index])) #acuracia
y_pred_mlp = clf.predict(X[test_index])
f1Score.append(f1_score(y[test_index], y_pred_mlp, average='binary'))
print('acurácia: ', fmean(score), " desvio padrão: ", (stdev(score)))
print('f1-score ', fmean(f1Score), "desvio padrão: ", (stdev(f1Score)))
file.write("\ncamadas: {}\tneuronios: {}\n".format(len(config_mlp),
config_mlp[0]))
file.write("{}\t{}\n".format(fmean(score), stdev(score)))
file.write("{}\t{}\n\n".format(fmean(f1Score), stdev(f1Score)))
def iterate(params):
p = params
limits = [None, None]
data = {}
p['percent_diff'] = None
p['count'] = 0
while check_completion(p['percent_diff'], p['precision']) == False:
p['count'] += 1
results = calculateData(p)
data = results[0]
p = results[1]
p['percent_diff'] = p['Pnf'] / p['Pnm']
if p['percent_diff'] <= 1:
limits[0] = p['IC_x']
if limits[1] == None:
p['IC_x'] *= 2
else:
p['IC_x'] = s.fmean(limits)
else:
limits[1] = p['IC_x']
if limits[0] == None:
p['IC_x'] /= 2
else:
p['IC_x'] = s.fmean(limits)
pause(limits, data, **p)
return data, p
def getMetadataMedia(metadatas):
if metadatas['danceability'] is None or len(
metadatas['danceability']) == 0:
return metadatas
# Fazer sistema de rankeamento, pegar o TIPO da musica
# Desvio Padrão
metadata = {}
metadata['danceability'] = "{:.3f}".format(
statistics.fmean(metadatas['danceability']))
metadata['energy'] = "{:.3f}".format(statistics.fmean(metadatas['energy']))
metadata['loudness'] = "{:.3f}".format(
statistics.fmean(metadatas['loudness']))
metadata['mode'] = statistics.mode(metadatas['mode'])
metadata['speechiness'] = "{:.3f}".format(
statistics.fmean(metadatas['speechiness']))
metadata['acousticness'] = "{:.3f}".format(
statistics.fmean(metadatas['acousticness']))
metadata['instrumentalness'] = "{:.3f}".format(
statistics.fmean(metadatas['instrumentalness']))
metadata['liveness'] = "{:.3f}".format(
statistics.fmean(metadatas['liveness']))
metadata['valence'] = "{:.3f}".format(
statistics.fmean(metadatas['valence']))
metadata['tempo'] = "{:.3f}".format(statistics.fmean(metadatas['tempo']))
metadata['duration_ms'] = "{:.3f}".format(
statistics.fmean(metadatas['duration_ms']))
return metadata
def log_evaluation(self, population: 'Population', **kwargs):
count_penalization = 0
partitions = list()
fitnesses = list()
costs = list()
generation = population.generation
for individual in population:
qtd = individual.qtd_partitions
if qtd > 1:
count_penalization += 1
partitions.append(qtd)
fitnesses.append(individual.fitness)
costs.append(individual.cost)
currentbest = population.current_best
bestoverall = population.documented_best
self.log("evaluation", generation, fmean(costs), pvariance(costs),
fmean(fitnesses), pvariance(fitnesses), count_penalization,
fmean(partitions), pvariance(partitions), currentbest.cost,
currentbest.fitness, bestoverall.cost, bestoverall.fitness,
bestoverall.last_improvement)
def __str__(self):
temp = fmean([fmean(i) for i in self.grades.values()
]) if self.grades.values() else 0
return (f"Имя: {self.name}\n"
f"Фамилия: {self.surname}\n"
f"Средняя оценка за домашние задания: {temp}\n"
f"Курсы в процессе изучения: {self.courses_in_progress}\n"
f"Завершенные курсы: {self.finished_courses}")
def __init__(self, DIR):
self.DIR = DIR
self.edges = purr.readXMLtag(purr.mrf(DIR, r"*.edg.xml"), "edge")
self.nodes = purr.readXMLtag(purr.mrf(DIR, r"*.nod.xml"), "node")
x = [float(node['x']) for node in self.nodes]
y = [float(node['y']) for node in self.nodes]
self.centroid = (stats.fmean(x), stats.fmean(y))
self.xMax = max(x)
self.xMin = min(x)
self.yMax = max(y)
self.yMin = min(y)
return
def analyze_clusters_intra_distances(evaluation, table_name):
clusters_mean_intra_distances = []
clusters_stdev_intra_distances = []
for cluster in evaluation.clusters.items():
clusters_mean_intra_distances.append(cluster[1].intra_dist_mean)
clusters_stdev_intra_distances.append(cluster[1].intra_dist_stdev)
evaluation.clusters_intra_dist_mean_mean = statistics.fmean(
clusters_mean_intra_distances)
evaluation.clusters_intra_dist_mean_stdev = statistics.stdev(
clusters_mean_intra_distances)
evaluation.clusters_intra_dist_stdev_mean = statistics.fmean(
clusters_stdev_intra_distances)
evaluation.clusters_intra_dist_stdev_stdev = statistics.stdev(
clusters_stdev_intra_distances)
def _iavg(*data: _num.Number) -> float:
"""
Intuitive AVeraGe
Returns the best average for the data set
Use fraction.Fraction instances for ratios
Ignores mode; nobody likes it
>>> iavg(3, 4, 5)
4
>>> from fractions import Fraction as F
>>> iavg(F(60, 20), F(200, 50), F(5, 1))
Fraction(180, 47)
"""
if len(data) == 1 and isinstance(data[0], (list, tuple)):
data = data[0]
target = 1.5 * iqr(*data)
q = qdict(*data)
for item in data:
if item < q['q1'] - target or item > q['q3'] + target:
return stat.median(data)
elif isinstance(item, fraction.Fraction):
return stat.harmonic_mean(data)
else:
return stat.fmean(data)
def measure_transitive_closure(iterations_num):
adj = timeit.repeat(
"Utils.get_transitive_closure_adj_matrix(intersection)",
setup="from __main__ import Utils, intersection",
repeat=iterations_num,
number=1)
sqr = timeit.repeat(
"Utils.get_transitive_closure_squaring(intersection)",
setup="from __main__ import Utils, intersection",
repeat=iterations_num,
number=1)
average_adj = round(fmean(adj), 6)
average_sqr = round(fmean(sqr), 6)
result = str(average_adj) + ' ' + str(average_sqr)
return result
def calculate_stats(self, data):
dates = [ d.get('date') for d in data ]
digital_values = [ d.get('digital_value') for d in data ]
timestamps = [ d.get('timestamp') for d in data ]
init_date, end_date = min(dates), max(dates)
init_timestamp, end_timestamp = min(timestamps), max(timestamps)
time_span = end_timestamp - init_timestamp
mean_value = fmean(digital_values)
n_samples = len(data)
sensor = data[0].get('sensor')
username = data[0].get('username')
ip = data[0].get('ip')
measure = data[0].get('measure')
return {
'digital_values': digital_values,
'end_date': end_date,
'end_timestamp': end_timestamp,
'init_date': init_date,
'init_timestamp': init_timestamp,
'ip': ip,
'mean_value': round(mean_value, 1),
'measure': measure,
'n_samples': n_samples,
'sensor': sensor,
'time_span': time_span,
'username': username
}
def statistics_element(self, element):
"""
Calcularemos el promedio.
Desde values obtendremos la informacion por elemento.
Usaremos 3 metodos: default mean, float mean y calculo manual.
"""
self.response = None
self.values = self.values_page(element)
if isinstance(self.values.get("data"), dict):
self.values_data = self.values.get("data").get("values")
self.values_list = [
value for value in self.values_data.values()
]
self.lenofvalues = len(self.values_list)
self.sumofvalues = sum(self.values_list)
self.meanofvalues = fmean(self.values_list)
self.fmeanofvalues = mean(self.values_list)
self.avgcomputofvalues = float(self.sumofvalues) / self.lenofvalues
self.response["data"] = {
"avg_stat-dmn": self.meanofvalues,
"avg_stat-fmn": self.fmeanofvalues,
"avg_comp-ute": self.avgcomputofvalues,
"sum": self.sumofvalues,
"len": self.lenofvalues
}
return self.response
def fmean(angka):
inp = angka.split(" ")
merge = []
for x in inp:
merge.append(int(x))
inp = statistics.fmean(merge)
print(inp)
def measure_print(iterations_num):
pr = timeit.repeat("Utils.print_label_matrices(intersection)",
setup="from __main__ import Utils, intersection",
repeat=iterations_num,
number=1)
average = round(fmean(pr), 6)
result = str(average)
return result
def calc(first, second, third, fourth):
"""Returns the average meaning from the list
and max argument."""
my_list = [first, second, third, fourth]
global MAX_NUMBER
if max(my_list) > MAX_NUMBER:
MAX_NUMBER = max(my_list)
average = statistics.fmean(my_list)
return average, MAX_NUMBER
def compute_averages(dataMatrix):
average = []
for classNum in range(len(dataMatrix)):
local_average = []
for i in (np.transpose(dataMatrix[classNum]))[:-1]:
mean = stat.fmean(i)
local_average.append(mean)
average.append(local_average)
return average
def seal_file(self):
date = self.current_date
i = 1
total_list = self.values[self.headline[1]]
count = len(total_list)
count_monotonic = 0
while i < count:
if total_list[i] >= total_list[i - 1]:
count_monotonic += 1
i += 1
result = {
"test": {
"name": "ddl"
},
"monotonicity": count_monotonic / len(total_list),
"size": {
"count": count,
"collectionCount": self.collections_per_database,
"indexesPercollection": self.indexes_per_collection,
"numberOfShards": self.no_shards,
"replicationFactor": self.replication_factor
},
"configuration": {
"version": self.version,
"branch": self.branch,
"mode": self.mode,
"edition": self.edition
},
"isoDate": date.isoformat(),
"date": date.timestamp(),
"ms": 1000 * date.timestamp()
}
i = 1
while i < len(self.headline):
name = self.headline[i]
test_values = self.values[name]
result['test'][name] = {
"average": statistics.fmean(test_values),
"median": statistics.median(test_values),
"min": min(test_values),
"max": max(test_values),
"deviation": {
"pst": statistics.pstdev(test_values),
"pvariance": statistics.pvariance(test_values),
"stdev": statistics.stdev(test_values),
"variance": statistics.variance(test_values)
},
"values": test_values,
"numberRuns": count
}
i += 1
print(json.dumps(result))
def stats(utt_len):
moy = stat.fmean(utt_len)
median = stat.median(utt_len)
var = stat.variance(utt_len)
quart1 = np.percentile(utt_len, 25)
quart3 = np.percentile(utt_len, 75)
printStats(moy, median, var, quart1, quart3)
return moy, median, var, quart1, quart3
def measure_intersection(iterations_num):
intersection = timeit.repeat(
"Utils.get_intersection(graph, automaton)",
setup="from __main__ import Utils, graph, automaton",
repeat=iterations_num,
number=1)
average = round(fmean(intersection), 6)
result = str(average)
return result
def main():
try:
data = json.loads(request.data)
validator.validate(data)
except ValidationError:
return jsonify(error=True, msg="JSON data invalid")
except:
return jsonify(error=True, msg="Invalid POST data")
query_types = list(data['query_type'].split('|'))
returns = {}
filtered_data = df
for _filter in data.get('filters', []):
if _filter['operand1'] == 'discount':
if not isinstance(_filter['operand2'], Number):
return jsonify(error=True,
msg="Discount percentage is not a number")
operator = _filter['operator']
percent = _filter['operand2']
filtered_data = filtered_data[filtered_data.price.apply(
filter_by_discount(operator, percent))]
elif _filter['operand1'] == 'brand.name':
operator = _filter['operator']
if operator != '==':
return jsonify(
error=True,
msg="To filter by brand name, use operator '=='")
brand_name = _filter['operand2']
filtered_data = filtered_data[filtered_data.brand.apply(
filter_by_brand(brand_name))]
if "discounted_products_list" in query_types:
returns.update({
'discounted_products_list': [x['$oid'] for x in filtered_data._id]
})
if "avg_discount" in query_types:
discounts = filtered_data.price.apply(discount_percentage)
avg_discount = fmean(discounts) if len(discounts) > 0 else 0.0
returns.update({'avg_discount': avg_discount})
if "expensive_list" in query_types:
expensive_list = filtered_data[filtered_data.apply(
filter_by_expensive_product, axis=1)]
returns.update(
{'expensive_list': [x['$oid'] for x in expensive_list._id]})
return jsonify(returns)
def get_average(
self,
value,
params=None): # returns the average value of a data header set
params = params or dict()
if value == 'smoker':
raise Exception("Boolean value is not valid entry")
elif value == 'sex' or value == 'region':
raise Exception("String value is not valid entry")
return round(statistics.fmean(self.get_data(value, params)), 2)
def get_celery_success(core_id, time_elap, fps_batch, wins, seeker_plot, hiding_plot ):
return {
'core_id': core_id,
'time_elapsed': time_elap,
'fps_peak': round(max(fps_batch)),
'fps_lower': round(min(fps_batch)),
'fps_mean': round(statistics.fmean(fps_batch)),
'fps_median': round(statistics.median(fps_batch)),
'fps_quantiles': [round(quantile) for quantile in statistics.quantiles(fps_batch)],
'wins': wins,
'seeker_plot': seeker_plot,
'hiding_plot': hiding_plot,
}
def measure_intra_distances(clusters, table_name):
for cluster_id in clusters:
cluster = clusters[cluster_id]
result = DB.exec(f"select Distance from {table_name} "
f"where Cluster = {cluster_id}")
count = 0
intra_distances = []
for row in result:
count += 1
intra_distances.append(row["Distance"])
cluster.intra_dist_mean = statistics.fmean(intra_distances)
cluster.intra_dist_stdev = statistics.stdev(
intra_distances) if count >= 2 else 0.0
def run_benchmarks(query, repeat=5):
results = defaultdict(list)
with get_new_connection() as connection:
for query in queries:
for _ in range(repeat):
start = time.perf_counter()
run_query_on_connection(connection, query)
results[query].append(time.perf_counter() - start)
return {
key: round(statistics.fmean(val), PRECISION)
for key, val in results.items()
}
def add_trade(self, time, price: float, volume: float, sell=False):
time = time_as_nano(time)
if time > self.end_time:
return False
trade_cost = price * volume
self.trade_cnt += 1
self.prices.append(price)
self.amounts.append(volume)
self.amounts.sort()
self.totals.append(trade_cost)
self.totals.sort()
if self.open == 0 or time < self.last_time:
self.open = price
if self.close == 0 or time > self.last_time:
self.close = price
if self.high == 0 or price > self.high:
self.high = price
if self.low == 0 or price < self.low:
self.low = price
self.volume += volume
if sell:
self.sell_totals.append(trade_cost)
self.sell_avg = statistics.fmean(self.sell_totals)
else:
self.buy_totals.append(trade_cost)
self.buy_avg = statistics.fmean(self.buy_totals)
# calc totals
self.trade_med = statistics.median(self.totals)
self.trade_avg = statistics.fmean(self.totals)
self.price_avg = statistics.fmean(self.prices)
self.last_time = time
def summary(self):
summary = {}
for section, durations in self.sections_duration.items():
p80, p95 = self.percentiles(durations)
summary[section] = {
'count': len(durations),
'min': round(min(durations) * 1000),
'max': round(max(durations) * 1000),
'avg': round(statistics.fmean(durations) * 1000),
'median': round(statistics.median(durations) * 1000),
'p80': round(p80 * 1000),
'p95': round(p95 * 1000),
}
return summary
def summary(self):
summary = {}
for section, durations in self.sections_duration.items():
p80, p95 = self.percentiles(durations)
summary[section] = {
"count": len(durations),
"min": round(min(durations) * 1000),
"max": round(max(durations) * 1000),
"avg": round(statistics.fmean(durations) * 1000),
"median": round(statistics.median(durations) * 1000),
"p80": round(p80 * 1000),
"p95": round(p95 * 1000),
}
return summary
def nayve_bayes_new(X, y, file):
clf = MultinomialNB()
skf = StratifiedKFold(n_splits=10)
skf.get_n_splits(X, y)
score = []
f1Score = []
for train_index, test_index in skf.split(X, y):
#print("TRAIN:", train_index, "TEST:", test_index)
score.append(
clf.fit(X[train_index],
y[train_index]).score(X[test_index],
y[test_index])) #acuracia
y_pred_mlp = clf.predict(X[test_index])
f1Score.append(f1_score(y[test_index], y_pred_mlp, average='binary'))
print('acurácia: ', fmean(score), " desvio padrão: ", stdev(score))
print('f1-score ', fmean(f1Score), "desvio padrão: ", stdev(f1Score))
file.write("{}\t{}\n".format(fmean(score), stdev(score)))
file.write("{}\t{}\n\n".format(fmean(f1Score), stdev(f1Score)))
def process_screen_rolling_backtest_result(json: dict, start_dt, end_dt,
precision):
if precision is None:
precision = 2
length = len(json['rows'])
data = [start_dt, end_dt, length]
data.extend(val for val in json['average'][4:8])
data.append(min(float(item[8]) for item in json['rows']))
data.append(max(float(item[9]) for item in json['rows']))
data.append(json['average'][10])
data.append(
round(statistics.fmean(float(item[5]) for item in json['rows'][0:13]
), precision) if length >= 13 else None)
data.append(
round(statistics.fmean(float(item[5]) for item in json['rows'][0:65]
), precision) if length >= 65 else None)
data.append(
round((statistics.geometric_mean(
float(item[5]) / 100 + 1 for item in json['rows']) - 1) * 100,
precision))
data.append(
round((statistics.geometric_mean(
float(item[6]) / 100 + 1 for item in json['rows']) - 1) * 100,
precision))
data.append(
round((statistics.geometric_mean(
float(item[5]) / 100 + 1 for item in json['rows'][0:13]) - 1) *
100, precision) if length >= 13 else None)
data.append(
round((statistics.geometric_mean(
float(item[5]) / 100 + 1 for item in json['rows'][0:65]) - 1) *
100, precision) if length >= 65 else None)
data.append(
round(statistics.fmean(float(item[4]) for item in json['rows'][0:65]
), precision) if length >= 65 else None)
return data
def main():
array = list(np.random.randint(low=1, high=50, size=10))
big_array = list(np.random.randint(low=1, high=100000, size=5000))
rprint("[green]Unsorted array:", array)
bubblesort(array)
rprint("[green]Sorted array:", array)
times = np.array([])
for i in range(10):
start_time = time.time()
bubblesort(big_array)
stop_time = time.time()
times = np.append(times, [stop_time - start_time])
rprint(
"[yellow]Average Time taken over 10 sorting operations to sort an \
array of size 5000:", fmean(times), "[yellow]seconds.")
def main():
random.seed(42)
array = [random.randint(a=1, b=50) for _ in range(10)]
big_array = [random.randint(a=1, b=100000) for _ in range(5000)]
rprint("[green]Unsorted array:", array)
radixsort(array)
rprint("[green]Sorted array:", array)
times = []
for i in range(10):
start_time = time.time()
radixsort(big_array)
stop_time = time.time()
times.append(stop_time - start_time)
rprint(
"[yellow]Average Time taken over 10 sorting operations to \
sort an array of size 5000:", fmean(times), "[yellow]seconds.")