def forecasts(request, watershed_dataset):
"""
Controller for forecasts
"""
# Create more user friendly names from the watershed_dataset variable
parts = watershed_dataset.split('-')
pretty_watershed_name = parts[1].replace('_', ' ').title()
pretty_subbasin_name = parts[2].replace('_', ' ').title()
title = pretty_watershed_name + ', ' + pretty_subbasin_name
# User friendly time
time = parts[-1]
year = time[:4]
month = time[4:6]
day = time[6:8]
hour = time[9:11]
pretty_time = '{0}/{1}/{2} @ {3}'.format(month, day, year, hour)
# Get a CKAN engine
ckan_engine = get_dataset_engine(name='nfie')
# Retrieve the dataset
response = ckan_engine.get_dataset(dataset_id=watershed_dataset)
# Assemble a list of buttons
buttons = []
if response['success']:
dataset = response['result']
for resource in dataset['resources']:
plot_url = reverse('ckan_app:plot', args=[resource['id']])
button = Button(display_text=resource['name'],
href=plot_url)
buttons.append(button)
context = {'title': title,
'pretty_time': pretty_time,
'buttons': buttons}
return render(request, 'ckan_app/forecasts.html', context)
def home(request):
"""
Controller for the app home page.
"""
# Get an engine object to work with CKAN
ckan_engine = get_dataset_engine(name="nfie")
# Get a list of datasets
response = ckan_engine.list_datasets()
# Assemble a list of ERFP datasets
erfp_dataset_names = []
if response['success']:
for dataset_name in response['result']:
if 'erfp' in dataset_name:
erfp_dataset_names.append(dataset_name)
# Assemble a list of unique watersheds
unique_watersheds = []
for erfp_dataset_name in erfp_dataset_names:
parts = erfp_dataset_name.split('-')
no_time = '-'.join(parts[:3])
if no_time not in unique_watersheds:
unique_watersheds.append(no_time)
# Assemble the list of options for the watershed select input
watershed_options = []
for watershed in unique_watersheds:
parts = watershed.split('-')
pretty_watershed_name = parts[1].replace('_', ' ').title()
pretty_subbasin_name = parts[2].replace('_', ' ').title()
combined = pretty_watershed_name + ', ' + pretty_subbasin_name
watershed_options.append((combined, watershed))
# Define the watershed select input
select_options = SelectInput(display_text='Watershed',
name='watershed',
multiple=False,
options=watershed_options)
# Assemble options for the time select input
time_options = []
show_time = False
selected_watershed = ''
# Handle form submission with the watershed select input
if request.POST and 'watershed' in request.POST:
selected_watershed = request.POST['watershed']
show_time = True
for dataset_name in erfp_dataset_names:
if selected_watershed in dataset_name:
parts = dataset_name.split('-')
time = parts[-1]
year = time[:4]
month = time[4:6]
day = time[6:8]
hour = time[9:11]
pretty_time = '{0}/{1}/{2} @ {3}'.format(month, day, year, hour)
time_options.append((pretty_time, dataset_name))
# Define time select input
time_select_options = SelectInput(display_text='Time',
name='time',
multiple=False,
options=time_options)
# Handle form submission with the time select input
if request.POST and 'time' in request.POST:
selected_dataset = request.POST['time']
return redirect('ckan_app:forecasts', watershed_dataset=selected_dataset)
context = {'unique_watersheds': unique_watersheds,
'select_options': select_options,
'time_select_options': time_select_options,
'show_time': show_time,
'selected_watershed': selected_watershed}
return render(request, 'ckan_app/home.html', context)
def plot(request, resource_id):
"""
Controller for forecast plots
"""
ckan_engine = get_dataset_engine(name='nfie')
response = ckan_engine.get_resource(resource_id=resource_id)
title = 'No Forecast Found'
pretty_time = ''
forecast_number = ''
timeseries = []
if response['success']:
resource = response['result']
# Get pretty watershed name
resource_name = resource['name']
parts = resource_name.split('-')
pretty_watershed_name = parts[1].replace('_', ' ').title()
pretty_subbasin_name = parts[2].replace('_', ' ').title()
title = pretty_watershed_name + ', ' + pretty_subbasin_name
# Get pretty time
time = parts[-2]
year = time[:4]
month = time[4:6]
day = time[6:8]
hour = time[9:11]
pretty_time = '{0}/{1}/{2} @ {3}'.format(month, day, year, hour)
# Get the forecast number
forecast_number = parts[-1]
# Create a workspace directory for the current user
current_username = request.user.username
controllers_directory = os.path.dirname(__file__)
user_workspace = os.path.join(controllers_directory, 'workspace', current_username)
if not os.path.exists(user_workspace):
os.makedirs(user_workspace)
# Clean out the workspace directory for the current user
for filename in os.listdir(user_workspace):
os.remove(os.path.join(user_workspace, filename))
# Download the resource into the current user workspace
file_name = os.path.join(user_workspace, 'temp.gz')
urllib.urlretrieve(resource['url'], file_name)
# Extract the archive file
with tarfile.open(file_name) as tar:
tar.extractall(user_workspace)
# Search directory for files with the netCDF extension (.nc)
netcdf_file_path = ''
for filename in os.listdir(user_workspace):
if '.nc' in filename:
netcdf_file_path = os.path.join(user_workspace, filename)
# Break the loop after the first one is found
# (hopefully, the directory has been cleared,
# so there should only be one)
break
# Open the netCDF file
data_nc = nc.Dataset(netcdf_file_path, mode="r")
# Arbitrary reach id
reach_index = 0
# Check the order of the time and comid fields and get the data vaules
qout_dimensions = data_nc.variables['Qout'].dimensions
if qout_dimensions[0].lower() == 'time' and qout_dimensions[1].lower() == 'comid':
data_values = data_nc.variables['Qout'][:,reach_index]
elif qout_dimensions[0].lower() == 'comid' and qout_dimensions[1].lower() == 'time':
data_values = data_nc.variables['Qout'][reach_index,:]
# Assemble the time stamps for each value given
# the starting date and a 6 hour timestep
start_date = datetime.datetime(int(year), int(month), int(day), int(hour))
timeseries = []
step = 0
for data_value in data_values.tolist():
hours = 6 * step
time = start_date + datetime.timedelta(hours=hours)
timeseries.append([time, data_value])
step += 1
# Close the netCDF file
data_nc.close()
# Define the time series plot
time_series_plot_object = HighChartsTimeSeries(
title='Streamflow Forecast',
y_axis_title='Streamflow',
y_axis_units='cms',
series=[{
'name': 'Streamflow',
'data': timeseries
}]
)
# Define the plot view
time_series_plot = PlotView(highcharts_object=time_series_plot_object,
width='100%',
height='500px')
context = {'title': title,
'pretty_time': pretty_time,
'forecast_number': forecast_number,
'time_series_plot': time_series_plot}
return render(request, 'ckan_app/plot.html', context)
def plot(request, resource_id):
"""
Controller for forecast plots
"""
ckan_engine = get_dataset_engine(name='nfie')
response = ckan_engine.get_resource(resource_id=resource_id)
resource = ''
timeseries = []
if response['success']:
resource = response['result']
resource_name = resource['name']
parts = resource_name.split('-')
pretty_watershed_name = parts[1].replace('_', ' ').title()
pretty_subbasin_name = parts[2].replace('_', ' ').title()
title = pretty_watershed_name + ', ' + pretty_subbasin_name
time = parts[-2]
year = time[:4]
month = time[4:6]
day = time[6:8]
hour = time[9:11]
pretty_time = '{0}/{1}/{2} @ {3}'.format(month, day, year, hour)
forecast_number = parts[-1]
current_username = request.user.username
contollers_directory = os.path.dirname(__file__)
user_workspace = os.path.join(contollers_directory, 'workspace', current_username)
if not os.path.exists(user_workspace):
os.makedirs(user_workspace)
for filename in os.listdir(user_workspace):
os.remove(os.path.join(user_workspace, filename))
file_name = os.path.join(user_workspace, 'temp.gz')
urllib.urlretrieve(resource['url'], file_name)
with tarfile.open(file_name) as tar:
tar.extractall(user_workspace)
netcdf_file_path = ''
for filename in os.listdir(user_workspace):
if '.nc' in filename:
netcdf_file_path = os.path.join(user_workspace, filename)
break
data_nc = nc.Dataset(netcdf_file_path, mode="r")
reach_index = 0
qout_dimensions = data_nc.variables['Qout'].dimensions
if qout_dimensions[0].lower() == 'time' and qout_dimensions[1].lower() == 'comid':
data_values = data_nc.variables['Qout'][:,reach_index]
elif qout_dimensions[0].lower() == 'comid' and qout_dimensions[1].lower() == 'time':
data_values = data_nc.variables['Qout'][reach_index,:]
start_date = datetime.datetime(int(year), int(month), int(day), int(hour))
step = 0
for data_value in data_values.tolist():
hours = 6 * step
time = start_date + datetime.timedelta(hours=hours)
timeseries.append([time, data_value])
step = step + 1
data_nc.close()
time_series_plot_object = HighChartsTimeSeries(
title='Streamflow Forecast',
y_axis_title='Streamflow',
y_axis_units='cms',
series=[{
'name': 'Streamflow',
'data': timeseries
}]
)
time_series_plot = PlotView(highcharts_object=time_series_plot_object,
width='100%',
height='500px')
context = {'title': title,
'pretty_time': pretty_time,
'forecast_number': forecast_number,
'time_series_plot': time_series_plot}
return render(request, 'ckan_app/plot.html', context)
