def get_all_points(self, range=None, use_file=False, point_threshold=None):
"""
get all points of every time series. Point is a date-time that the time-series has data for
returns a dictionary of the form {"time-series name": np.array([d1,d2,d3, ...])}
dx are of the form "month/day/year-hours:minutes:seconds"
dx are ordered
range is used to filter the time series whose points will be returned
range = [start_date, end_date]
"""
if not use_file or not os.path.exists("all-date-time-points"):
self.assert_connected()
tsnames_list = self.get_distinct_names(range=range, point_threshold=point_threshold)
dt = {}
for name in tsnames_list:
c = self.execute_query("select name, date, time from dataset "
"where name='%s' order by date, time" % name)
# d = self.db_name.execute_query("select count(date) from dataset "
# "where name='%s'" % name).fetchone()[0]
# d = int(d)
for res in c:
name = res[0]
date_time = res[1] + "-" + res[2]
if name in dt:
dt[name].append(date_time)
else:
dt[name] = [date_time]
# print(d, len(dt[name]))
# assert d == len(dt[name])
if use_file:
with open("all-date-time-points", 'w') as f:
for key, value in dt.items():
print(key + "," + ",".join(value), file=f)
else:
dt = {}
with open("all-date-time-points", 'r') as f:
for line in f:
line = line[:-1]
split_line = line.split(",")
dt[split_line[0]] = split_line[1:]
for key, value in dt.items():
dt[key] = np.array(value)
return dt
def sortedDict(d):
temp = sorted(d.items(), key=lambda x: x[1], reverse=True)
sorted_dict = {}
for item in temp:
sorted_dict[item[0]] = item[1]
return sorted_dict
def maxDictVal(dt):
'''
Finds the key with the max integer value inside a dictionary and returns it
Note: found on Stackoverflow - https://stackoverflow.com/questions/268272/getting-key-with-maximum-value-in-dictionary
'''
try:
topKey = max(dt.items(), key=operator.itemgetter(1))[0]
except:
topKey = -1
return topKey
def print_max_date_times(self):
"""
print the max (end) date-time of every time-series
"""
self.assert_connected()
tsnames_list = self.get_distinct_names()
dt = {}
for name in tsnames_list:
date_time = self.get_last_datetime(name)
if date_time in dt:
dt[date_time] += 1
continue
else:
dt[date_time] = 1
for key, value in sorted(dt.items()):
print(key + " " + str(value))
def get_start_end_points(self, range=None, use_file=False, point_threshold=None):
"""
get the start date-time and end date-time of every time-series
returns a dictionary of the form {"time-series name": [start_datetime, end_datetime]}
start_datetime and end_datetime are of the form "month/day/year-hours:minutes:seconds"
range is used to filter the time series whose points will be returned
range = [start_date, end_date]
"""
if not use_file or not os.path.exists("date-time-pairs"):
self.assert_connected()
tsnames_list = self.get_distinct_names(range=range, point_threshold=point_threshold)
dt = {}
for name in tsnames_list:
c = self.execute_query("select name, min(date), min(time), max(date), max(time) from dataset "
"where name='%s'" % name)
res = c.fetchone()
name = res[0]
min_date_time = res[1] + "-" + res[2]
max_date_time = res[3] + "-" + res[4]
dt[name] = [min_date_time, max_date_time]
if use_file:
with open("date-time-pairs", 'w') as f:
for key, value in dt.items():
print(key + "," + value[0] + "," + value[1], file=f)
else:
dt = {}
with open("date-time-pairs", 'r') as f:
for line in f:
line = line[:-1]
split_line = line.split(",")
dt[split_line[0]] = [split_line[1], split_line[2]]
return dt
def invertDictionary(dt):
return dict([[v, k] for k, v in dt.items()])
