def get_live_report(cls, entities):
base_directory = cls.get_entity_base_directory()
report = {}
live_headers = cls.live_csv_headers if cls.live_csv_headers else cls.csv_headers
for index, header in enumerate(live_headers):
report[header] = copy.deepcopy(cls.csv_default_values[index])
times = []
live_report_paths = []
for entity in entities:
entity_directory = os.path.join(base_directory, entity)
reports_directory = os.path.join(entity_directory, "reports",
cls.reports_folder)
file_path = os.path.join(reports_directory, "live.csv")
if not os.path.exists(file_path):
continue
live_report_paths.append(file_path)
with open(file_path, "r") as live_file:
lastest_entry = deque(csv.DictReader(live_file), 1)[0]
times.append(
datetime.strptime(lastest_entry["Time"],
"%Y-%m-%d %H:%M:%S"))
for header in live_headers:
if lastest_entry[header][0].isdigit():
report[header] += int(
ast.literal_eval(lastest_entry[header]))
else: # It's a list
entry = ast.literal_eval(lastest_entry[header])
if is_list_recursively_empty(report[header]):
report[header] = entry
else:
report[header] = [
a + b for a, b in zip(report[header], entry)
]
report["Time"] = ""
report["Trend"] = 0
if times:
report["Time"] = str(min(times))
trend_live_values = cls.get_trend_live_values(live_report_paths)
if trend_live_values:
report["Trend"] = cls.calculate_trend_value(trend_live_values)
return report
def get_hourly_report(cls, entities: List[str], report_date: date) -> Dict:
base_directory = cls.get_entity_base_directory()
hours = list(range(0, 24))
results = {}
for index, header in enumerate(cls.csv_headers):
if cls.csv_default_values[index] == 0:
results[header] = np.zeros(24)
else:
results[header] = []
for entity in entities:
entity_directory = os.path.join(base_directory, entity)
reports_directory = os.path.join(entity_directory, "reports",
cls.reports_folder)
file_path = os.path.join(reports_directory,
f"report_{report_date}.csv")
if os.path.exists(file_path):
df = pd.read_csv(file_path)
for header in cls.csv_headers:
if is_numeric_dtype(df[header]):
results[header] += np.pad(
df[header].to_numpy(),
(0, 24 - df[header].to_numpy().size),
mode="constant")
else: # It's a list
values = df[header].apply(ast.literal_eval).tolist()
entry = np.pad(values, (0, 24 - len(values)),
mode="constant").tolist()
if is_list_recursively_empty(results[header]):
results[header] = entry
else:
results[header] = [[
c + d for c, d in zip(a, b)
] for a, b in zip(results[header], entry)]
for metric in results:
results[metric] = list(results[metric])
results["Hours"] = hours
return results
def get_weekly_report(cls,
entities: List[str],
number_of_weeks: int = 0,
from_date: date = None,
to_date: date = None) -> Dict:
# The In/Out metric cannot be fully aggregated using "sum"
weekly_report_data = cls.generate_weekly_report_data(
entities, number_of_weeks, from_date, to_date)
report = {
"Weeks": [],
"InMax": [],
"OutMax": [],
"InAvg": [],
"OutAvg": [],
}
for header in cls.csv_headers:
report[header] = []
for week, week_data in weekly_report_data.items():
estimated_max_occ = max(
week_data["EstimatedMaxOccupancy"]
) if week_data["EstimatedMaxOccupancy"] else 0
estimated_avg_occ = round(
mean(week_data["EstimatedAverageOccupancy"]),
2) if week_data["EstimatedAverageOccupancy"] else 0
estimated_latest_occ = round(
week_data["EstimatedLatestOccupancy"]
[-1]) if week_data["EstimatedLatestOccupancy"] else 0
in_sum = sum(week_data["In"])
out_sum = sum(week_data["Out"])
in_max = max(week_data["In"]) if week_data["In"] else 0
out_max = max(week_data["Out"]) if week_data["Out"] else 0
in_avg = round(mean(week_data["In"]), 2) if week_data["In"] else 0
out_avg = round(mean(week_data["Out"]),
2) if week_data["Out"] else 0
report["Weeks"].append(week)
report["In"].append(in_sum)
report["Out"].append(out_sum)
report["InMax"].append(in_max)
report["OutMax"].append(out_max)
report["InAvg"].append(in_avg)
report["OutAvg"].append(out_avg)
report["EstimatedMaxOccupancy"].append(estimated_max_occ)
report["EstimatedAverageOccupancy"].append(estimated_avg_occ)
report["EstimatedLatestOccupancy"].append(estimated_latest_occ)
if is_list_recursively_empty(week_data["Summary"]):
boundary_name = []
weekly_in = []
weekly_out = []
else:
boundary_names, weekly_in, weekly_out = list(
zip(*week_data["Summary"]))
boundary_name = next(x for x in boundary_names
if not is_list_recursively_empty(x))
weekly_in = _fill_partially_empty_result(weekly_in, 0)
weekly_out = _fill_partially_empty_result(weekly_out, 0)
report["Summary"].append([
boundary_name, [sum(x) for x in zip(*weekly_in)],
[sum(x) for x in zip(*weekly_out)]
])
return report