def extract_phase(zdrdir, datadir, date, outdir):
filelist = glob.glob(datadir + date + '/*.nc')
if len(filelist) > 0:
ml_file = os.path.join(zdrdir, date, 'day_ml_zdr.csv')
ml_data = pd.read_csv(ml_file, index_col=0, parse_dates=True)
rain_index = np.where(ml_data['ML'].notnull())[0]
all_phase = []
time = []
for file in rain_index[0:10]:
print(filelist[file])
rad = pyart.io.read(filelist[file], delay_field_loading=True)
try:
IP = pyart.correct.phase_proc.det_sys_phase(
rad,
ncp_lev=0.4,
rhohv_lev=0.6,
ncp_field='SQI',
rhv_field='RhoHV',
phidp_field='uPhiDP')
all_phase.append(IP)
time_of_file = filelist[file][120:135]
time.append(time_of_file)
except:
print(date, 'no uphidp')
return False
pdtime = pd.to_datetime(time, format='%Y%m%d-%H%M%S')
data = pd.DataFrame({'Initial Phase': all_phase}, index=pdtime)
phase_file = os.path.join(outdir, 'initial_phase_' + date + '.csv')
data.to_csv(phase_file)
return True
def get_plot():
img = io.BytesIO()
time = []
consumed = []
produced = []
with open('test1.csv') as data:
next(data)
for line in csv.reader(data):
datetime_obj = datetime.strptime(line[0], '%d-%m-%y %H:%M')
time.append(datetime_obj)
consumed.append(line[1])
produced.append(line[2].strip('-'))
cons = go.Scatter(x=time,
y=consumed,
name="Consumed Electricity",
line=dict(color='#17BECF'),
opacity=0.8)
prod = go.Scatter(x=time,
y=produced,
name="Produced Electricity",
line=dict(color='#7F7F7F'),
opacity=0.8)
data = [cons, prod]
layout = go.Layout(title='Electricity consumption and production',
xaxis=dict(title='Time', ),
yaxis=dict(title='kWh', ))
fig = dict(data=data, layout=layout)
return plot(fig, output_type='div')
def cast():
if request.method == "GET":
return render_template("forecast.html")
else:
city = request.form.get("location")
result = weatherapp(city, None)
if result["cod"] == "404" or result["name"] == "None":
return apology("No data found for your search request", 403)
data = forecast(result["coord"])
timezoneoff = data["timezone_offset"]
sunrise = []
sunset = []
time = []
for info in data["daily"]:
sunrise.append(
datetime.fromtimestamp(info["sunrise"] + timezoneoff).time())
sunset.append(
datetime.fromtimestamp(info["sunset"] + timezoneoff).time())
for i in range(8):
time.append((dt.datetime.today() + dt.timedelta(days=i)).date())
return render_template("forecastResults.html",
data=data,
city=city,
sunrise=sunrise,
sunset=sunset,
time=time)
def get_full_datetime(self):
time = []
if self.time_start:
time.append(self.time_start.strftime('%H:%M'))
if self.time_end:
time.append(self.time_end.strftime('%H:%M'))
return '%s %s' % (self.date.strftime('%d.%m.%Y'), '-'.join(time))
def flight_duration_list(self, data):
""" This method retrieve a list which contains flight duration """
time = []
try:
for path, _, node in jxmlease.parse(
data,
generator=
"tir38:TravelItinerary/tir38:ItineraryInfo/tir38:ReservationItems/tir38:Item/tir38:FlightSegment"
):
time1 = node.get_xml_attr('ElapsedTime')
time.append(str(time1))
except:
time = ['N/A']
return time
region = df['Region'].value_counts()
city = df['City'].value_counts()
df.describe()
df.dtypes
from datetime import datetime, time
date = []
time = []
for i in range(0, 127826):
a = i
aa = df.timestamp[i][0:10]
date.append(aa)
ab = df.timestamp[i][11:-1]
time.append(ab)
date.append(df.timestamp[125825][0:10])
time.append(df.timestamp[125825][11:-1])
# Creating a Date column to store the actual Date format for the given Month column
df["Date"] = date
df["time"] = time
df["Date"] = pd.to_datetime(df.Date)
df["time"] = pd.to_datetime(df.time)
df["Date1"] = pd.to_datetime(df.Date, format="%b-%y")
# Extracting Day, weekday name, month name, year from the Date column using
# Date functions from pandas
def calculate_activity_level(cow):
# imports
import pandas as pd
from datetime import date
from datetime import time
from datetime import datetime
from datetime import timedelta
# convert time column from str to datetime type
# sort in ascending order (chronologically)
cow['time'] = pd.to_datetime(cow['time'])
cow = cow.sort_values(by=['time'])
# find the initial time value
starting_time = cow.loc[0, 'time']
# initialize the lists that are to be appended in the loop
time = []
activity_level = []
# create one hour time slices
end_time = starting_time
while (end_time <= cow.iloc[len(cow) - 1, 0]):
# create the 1-hour time range for the data to be filtered in
new_time = end_time
end_time = new_time + timedelta(hours=1) # one hour slice
# create date filter mask
# greater than the start date and smaller than the end date
# hold the data from the 1-hour timeslice in a placeholder dataframe
mask = (cow['time'] > new_time) & (cow['time'] <= end_time)
placeholder = cow[mask]
# summarise the value counts for the labels in the placeholder dataframe
summary = pd.DataFrame(placeholder['labels'].value_counts())
summary['cluster'] = summary.index
# rename columns
summary.rename(columns={summary.columns[0]: "count"}, inplace=True)
# add weights column
# tabulate the values in the weight column
# IMPORTANT: assumption is cluster 0 is low activity, cluster 1 is medium activity, cluster 2 is high activity
summary['weight'] = None
for i in range(len(summary)):
if (summary.iloc[i, 1] == 0):
summary.loc[i, 'weight'] = 0
elif (summary.iloc[i, 1] == 1):
summary.loc[i, 'weight'] = 0.1
else:
summary.loc[i, 'weight'] = 0.9
# initialise the hourly activity level as an empty list
hourly_activity_level = []
# keep appending the product of weight and value count for each cluster label to the hourly_activity_level list
for i in range(len(summary)):
activity_level.append(summary.iloc[i, 0] * summary.iloc[i, 2])
# calculate the sum of the elements in the hourly_activity_level list
hourly_activity_level = sum(hourly_activity_level)
# append the time and sum of hourly activity level to the time and activity_level lists
time.append(new_time)
activity_level.append(hourly_activity_level)
# create a dataframe using the time and activity level lists
activity_df = pd.DataFrame(list(zip(time, activity_level)),
columns=['time', 'activity_level'])
# convert the time column to datetime object
activity_df['time'] = pd.to_datetime(activity_df['time'])
# initialise the activity level columns for the previous 1, 24, 48, and 72 hours
activity_df['activity_level_1'] = None
activity_df['activity_level_24'] = None
activity_df['activity_level_48'] = None
activity_df['activity_level_72'] = None
for i in range(len(activity_df)):
timevalue_1 = activity_df.loc[i, 'time'] - timedelta(hours=1)
timevalue_24 = activity_df.loc[i, 'time'] - timedelta(hours=24)
timevalue_48 = activity_df.loc[i, 'time'] - timedelta(hours=48)
timevalue_72 = activity_df.loc[i, 'time'] - timedelta(hours=72)
# some errors arise due to duplicate time values being present in the data
# trying to cheat my way through with the use of try except block
try:
activity_df.loc[i, 'activity_level_1'] = activity_df.loc[
activity_df['time'] ==
timevalue_1]['activity_level'].values[0]
activity_df.loc[i, 'activity_level_24'] = activity_df.loc[
activity_df['time'] ==
timevalue_24]['activity_level'].values[0]
activity_df.loc[i, 'activity_level_48'] = activity_df.loc[
activity_df['time'] ==
timevalue_48]['activity_level'].values[0]
activity_df.loc[i, 'activity_level_72'] = activity_df.loc[
activity_df['time'] ==
timevalue_72]['activity_level'].values[0]
except:
pass
# return the activity level dataframe
return activity_df