def _prepare_datasets_for_evaluation(reference, targets, config_data):
""""""
subset = config_data['evaluation'].get('subset', None)
temporal_time_delta = config_data['evaluation'].get(
'temporal_time_delta', None)
spatial_regrid_lats = config_data['evaluation'].get(
'spatial_regrid_lats', None)
spatial_regrid_lons = config_data['evaluation'].get(
'spatial_regrid_lons', None)
# If we have a temporal time delta and it's daily (i.e., 1) we will
# normalize the data as daily data (which means we adjust the start times
# for each bucket of data to be consistent). By default we will normalize
# the data as monthly. Note that this will not break yearly data so it's
# safer to do this no matter what. This keeps us from ending up with 1-off
# errors in the resulting dataset shape post-temporal/spatial adjustments
# that break evaluations.
string_time_delta = 'monthly'
if temporal_time_delta and temporal_time_delta == 1:
string_time_delta = 'daily'
reference = dsp.normalize_dataset_datetimes(reference, string_time_delta)
targets = [
dsp.normalize_dataset_datetimes(t, string_time_delta) for t in targets
]
if subset:
start = dateutil.parser.parse(subset[4])
end = dateutil.parser.parse(subset[5])
bounds = Bounds(subset[0], subset[1], subset[2], subset[3], start, end)
if reference:
reference = dsp.safe_subset(bounds, reference)
if targets:
targets = [dsp.safe_subset(bounds, t) for t in targets]
if temporal_time_delta:
resolution = timedelta(temporal_time_delta)
if reference:
reference = dsp.temporal_rebin(reference, resolution)
if targets:
targets = [dsp.temporal_rebin(t, resolution) for t in targets]
if spatial_regrid_lats and spatial_regrid_lons:
lats = np.arange(spatial_regrid_lats[0], spatial_regrid_lats[1],
spatial_regrid_lats[2])
lons = np.arange(spatial_regrid_lons[0], spatial_regrid_lons[1],
spatial_regrid_lons[2])
if reference:
reference = dsp.spatial_regrid(reference, lats, lons)
if targets:
targets = [dsp.spatial_regrid(t, lats, lons) for t in targets]
return reference, targets
def _prepare_datasets_for_evaluation(reference, targets, config_data):
""""""
subset = config_data['evaluation'].get('subset', None)
temporal_time_delta = config_data['evaluation'].get('temporal_time_delta', None)
spatial_regrid_lats = config_data['evaluation'].get('spatial_regrid_lats', None)
spatial_regrid_lons = config_data['evaluation'].get('spatial_regrid_lons', None)
# If we have a temporal time delta and it's daily (i.e., 1) we will
# normalize the data as daily data (which means we adjust the start times
# for each bucket of data to be consistent). By default we will normalize
# the data as monthly. Note that this will not break yearly data so it's
# safer to do this no matter what. This keeps us from ending up with 1-off
# errors in the resulting dataset shape post-temporal/spatial adjustments
# that break evaluations.
string_time_delta = 'monthly'
if temporal_time_delta and temporal_time_delta == 1:
string_time_delta = 'daily'
reference = dsp.normalize_dataset_datetimes(reference, string_time_delta)
targets = [dsp.normalize_dataset_datetimes(t, string_time_delta) for t in targets]
if subset:
start = dateutil.parser.parse(subset[4])
end = dateutil.parser.parse(subset[5])
bounds = Bounds(subset[0], subset[1], subset[2], subset[3], start, end)
if reference:
reference = dsp.safe_subset(bounds, reference)
if targets:
targets = [dsp.safe_subset(bounds, t) for t in targets]
if temporal_time_delta:
resolution = timedelta(temporal_time_delta)
if reference:
reference = dsp.temporal_rebin(reference, resolution)
if targets:
targets = [dsp.temporal_rebin(t, resolution) for t in targets]
if spatial_regrid_lats and spatial_regrid_lons:
lats = np.arange(spatial_regrid_lats[0], spatial_regrid_lats[1], spatial_regrid_lats[2])
lons = np.arange(spatial_regrid_lons[0], spatial_regrid_lons[1], spatial_regrid_lons[2])
if reference:
reference = dsp.spatial_regrid(reference, lats, lons)
if targets:
targets = [dsp.spatial_regrid(t, lats, lons) for t in targets]
return reference, targets
def temporalRegrid(dataset, timeRes=timedelta(days=365)):
'''Temporally rebin a dataset variable to a specified time resolution (timedelta object).'''
dataset = dsp.temporal_rebin(dataset, timeRes)
name = dataset.name
if name is None: name = ''
print('temporalRebin: Dataset %s has new shape %s' % (name, str(dataset.values.shape)), file=sys.stderr)
return dataset
def check_some_dsp_functions(dataset):
'''
Run a subset of dataset processor functions and check for
any kind of exception.
'''
try:
dsp.temporal_rebin(dataset, 'annual')
dsp.ensemble([dataset])
except Exception as e:
fail("\nDataset processor functions")
print("Following error occured:")
print(str(e))
end()
finally:
os.remove(dataset.origin['path'])
success("\nDataset processor functions")
def check_some_dsp_functions(dataset):
'''
Run a subset of dataset processor functions and check for
any kind of exception.
'''
try:
dsp.temporal_rebin(dataset, 'annual')
dsp.ensemble([dataset])
except Exception as e:
fail("\nDataset processor functions")
print "Following error occured:"
print str(e)
end()
finally:
os.remove(dataset.origin['path'])
success("\nDataset processor functions")
def test_variable_propagation(self):
annual_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset,
datetime.timedelta(days=365))
self.assertEquals(annual_dataset.name,
self.ten_year_monthly_dataset.name)
self.assertEquals(annual_dataset.variable,
self.ten_year_monthly_dataset.variable)
def test_variable_propagation(self):
annual_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset,
"annual")
self.assertEquals(annual_dataset.name,
self.ten_year_monthly_dataset.name)
self.assertEquals(annual_dataset.variable,
self.ten_year_monthly_dataset.variable)
def test_daily_to_monthly_rebin(self):
"""This test takes a really long time to run. TODO: Figure out where the performance drag is"""
monthly_dataset = dp.temporal_rebin(self.two_years_daily_dataset, datetime.timedelta(days=31))
bins = list(set([datetime.datetime(time_reading.year, time_reading.month, 1) for time_reading in self.two_years_daily_dataset.times]))
bins = np.array(bins)
bins.sort()
np.testing.assert_array_equal(monthly_dataset.times, bins)
def test_non_rebin(self):
"""This will take a monthly dataset and ask for a monthly rebin of 28 days. The resulting
dataset should have the same time values"""
monthly_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset,
datetime.timedelta(days=28))
good_times = self.ten_year_monthly_dataset.times
np.testing.assert_array_equal(monthly_dataset.times, good_times)
def test_variable_propagation(self):
annual_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset,
datetime.timedelta(days=365))
self.assertEquals(annual_dataset.name,
self.ten_year_monthly_dataset.name)
self.assertEquals(annual_dataset.variable,
self.ten_year_monthly_dataset.variable)
def test_variable_propagation(self):
annual_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset,
"annual")
self.assertEquals(annual_dataset.name,
self.ten_year_monthly_dataset.name)
self.assertEquals(annual_dataset.variable,
self.ten_year_monthly_dataset.variable)
def temporalRegrid(dataset, timeRes=timedelta(days=365)):
'''Temporally rebin a dataset variable to a specified time resolution (timedelta object).'''
dataset = dsp.temporal_rebin(dataset, timeRes)
name = dataset.name
if name is None: name = ''
print >> sys.stderr, 'temporalRebin: Dataset %s has new shape %s' % (
name, str(dataset.values.shape))
return dataset
def test_non_rebin(self):
"""This will take a monthly dataset and ask for a monthly rebin of 28 days. The resulting
dataset should have the same time values"""
monthly_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset, "monthly")
bins = list(set([datetime.datetime(time_reading.year, time_reading.month, 15) for time_reading in self.ten_year_monthly_dataset.times]))
bins = np.array(bins)
bins.sort()
np.testing.assert_array_equal(monthly_dataset.times, bins)
def test_daily_to_annual_rebin(self):
annual_dataset = dp.temporal_rebin(
self.two_years_daily_dataset, "annual")
bins = list(set([datetime.datetime(
time_reading.year, 7, 2)
for time_reading in self.two_years_daily_dataset.times]))
bins = np.array(bins)
bins.sort()
np.testing.assert_array_equal(annual_dataset.times, bins)
def test_daily_to_annual_rebin(self):
annual_dataset = dp.temporal_rebin(
self.two_years_daily_dataset, "annual")
bins = list(set([datetime.datetime(
time_reading.year, 7, 2)
for time_reading in self.two_years_daily_dataset.times]))
bins = np.array(bins)
bins.sort()
np.testing.assert_array_equal(annual_dataset.times, bins)
def test_non_rebin(self):
"""
This will take a monthly dataset and ask for a monthly rebin of
28 days. The resulting dataset should have the same time values
"""
monthly_dataset = dp.temporal_rebin(
self.ten_year_monthly_dataset, "monthly")
bins = list(set([datetime.datetime(
time_reading.year, time_reading.month, 15)
for time_reading in self.ten_year_monthly_dataset.times]))
bins = np.array(bins)
bins.sort()
np.testing.assert_array_equal(monthly_dataset.times, bins)
def _prepare_datasets_for_evaluation(reference, targets, config_data):
""""""
subset = config_data['evaluation'].get('subset', None)
temporal_time_delta = config_data['evaluation'].get('temporal_time_delta', None)
spatial_regrid_lats = config_data['evaluation'].get('spatial_regrid_lats', None)
spatial_regrid_lons = config_data['evaluation'].get('spatial_regrid_lons', None)
if subset:
start = dateutil.parser.parse(subset[4])
end = dateutil.parser.parse(subset[5])
bounds = Bounds(subset[0], subset[1], subset[2], subset[3], start, end)
if reference:
reference = dsp.safe_subset(bounds, reference)
if targets:
targets = [dsp.safe_subset(bounds, t) for t in targets]
if temporal_time_delta:
resolution = timedelta(temporal_time_delta)
if reference:
reference = dsp.temporal_rebin(reference, resolution)
if targets:
targets = [dsp.temporal_rebin(t, resolution) for t in targets]
if spatial_regrid_lats and spatial_regrid_lons:
lats = np.arange(spatial_regrid_lats[0], spatial_regrid_lats[1], spatial_regrid_lats[2])
lons = np.arange(spatial_regrid_lons[0], spatial_regrid_lons[1], spatial_regrid_lons[2])
if reference:
reference = dsp.spatial_regrid(reference, lats, lons)
if targets:
targets = [dsp.spatial_regrid(t, lats, lons) for t in targets]
return reference, targets
cru31_dataset = rcmed.parameter_dataset(dataset_id,
parameter_id,
min_lat,
max_lat,
min_lon,
max_lon,
start_time,
end_time)
# Step 3: Resample Datasets so they are the same shape.
# Running Temporal Rebin early helps negate the issue of datasets being on different
# days of the month (1st vs. 15th)
print("Temporally Rebinning the Datasets to an Annual Timestep")
# To run annual temporal Rebinning,
knmi_dataset = dsp.temporal_rebin(knmi_dataset, temporal_resolution='annual')
dataset_start, dataset_end = knmi_dataset.temporal_boundaries()
start_time = max([start_time, dataset_start])
end_time = min([end_time, dataset_end])
wrf311_dataset = dsp.temporal_rebin(
wrf311_dataset, temporal_resolution='annual')
dataset_start, dataset_end = wrf311_dataset.temporal_boundaries()
start_time = max([start_time, dataset_start])
end_time = min([end_time, dataset_end])
cru31_dataset = dsp.temporal_rebin(cru31_dataset, temporal_resolution='annual')
dataset_start, dataset_end = cru31_dataset.temporal_boundaries()
start_time = max([start_time, dataset_start])
end_time = min([end_time, dataset_end])
SEASON_MONTH_END = 12
EVAL_BOUNDS = Bounds(LAT_MIN, LAT_MAX, LON_MIN, LON_MAX, START, END)
# Normalize the time values of our datasets so they fall on expected days
# of the month. For example, monthly data will be normalized so that:
# 15 Jan 2014, 15 Feb 2014 => 1 Jan 2014, 1 Feb 2014
ref_dataset = dsp.normalize_dataset_datetimes(ref_dataset, "monthly")
target_dataset = dsp.normalize_dataset_datetimes(target_dataset, "monthly")
# Subset down the evaluation datasets to our selected evaluation bounds.
target_dataset = dsp.subset(EVAL_BOUNDS, target_dataset)
ref_dataset = dsp.subset(EVAL_BOUNDS, ref_dataset)
# Do a monthly temporal rebin of the evaluation datasets.
target_dataset = dsp.temporal_rebin(target_dataset,
datetime.timedelta(days=30))
ref_dataset = dsp.temporal_rebin(ref_dataset, datetime.timedelta(days=30))
# Spatially regrid onto a 1 degree lat/lon grid within our evaluation bounds.
new_lats = np.arange(LAT_MIN, LAT_MAX, 1.0)
new_lons = np.arange(LON_MIN, LON_MAX, 1.0)
target_dataset = dsp.spatial_regrid(target_dataset, new_lats, new_lons)
ref_dataset = dsp.spatial_regrid(ref_dataset, new_lats, new_lons)
# Load the datasets for the evaluation.
mean_bias = metrics.MeanBias()
# These versions of the metrics require seasonal bounds prior to running
# the metrics. You should set these values above in the evaluation
# configuration section.
spatial_std_dev_ratio = metrics.SeasonalSpatialStdDevRatio(
month_start=SEASON_MONTH_START, month_end=SEASON_MONTH_END)
# With monthly data, we could have data falling on the 1st, 15th, or some other # day of the month. Let's fix that real quick. ################################################################################ knmi_dataset = dsp.normalize_dataset_datetimes(knmi_dataset, 'monthly') wrf_dataset = dsp.normalize_dataset_datetimes(wrf_dataset, 'monthly') # We're only going to run this evaluation over a years worth of data. We'll # make a Bounds object and use it to subset our datasets. ################################################################################ subset = Bounds(-45, 42, -24, 60, datetime.datetime(1989, 1, 1), datetime.datetime(1989, 12, 1)) knmi_dataset = dsp.subset(subset, knmi_dataset) wrf_dataset = dsp.subset(subset, wrf_dataset) # Temporally re-bin the data into a monthly timestep. ################################################################################ knmi_dataset = dsp.temporal_rebin(knmi_dataset, datetime.timedelta(days=30)) wrf_dataset = dsp.temporal_rebin(wrf_dataset, datetime.timedelta(days=30)) # Spatially regrid the datasets onto a 1 degree grid. ################################################################################ # Get the bounds of the reference dataset and use it to create a new # set of lat/lon values on a 1 degree step # Using the bounds we will create a new set of lats and lons on 1 degree step min_lat, max_lat, min_lon, max_lon = knmi_dataset.spatial_boundaries() new_lons = numpy.arange(min_lon, max_lon, 1) new_lats = numpy.arange(min_lat, max_lat, 1) # Spatially regrid datasets using the new_lats, new_lons numpy arrays knmi_dataset = dsp.spatial_regrid(knmi_dataset, new_lats, new_lons) wrf_dataset = dsp.spatial_regrid(wrf_dataset, new_lats, new_lons)
urlretrieve(FILE_LEADER + FILE_2, FILE_2_PATH)
# Step 1: Load Local NetCDF Files into OCW Dataset Objects.
print("Loading %s into an OCW Dataset Object" % (FILE_1_PATH,))
knmi_dataset = local.load_file(FILE_1_PATH, "tasmax")
print("KNMI_Dataset.values shape: (times, lats, lons) - %s \n" %
(knmi_dataset.values.shape,))
print("Loading %s into an OCW Dataset Object" % (FILE_2_PATH,))
wrf_dataset = local.load_file(FILE_2_PATH, "tasmax")
print("WRF_Dataset.values shape: (times, lats, lons) - %s \n" %
(wrf_dataset.values.shape,))
# Step 2: Temporally Rebin the Data into an Annual Timestep.
print("Temporally Rebinning the Datasets to an Annual Timestep")
knmi_dataset = dsp.temporal_rebin(knmi_dataset, temporal_resolution='annual')
wrf_dataset = dsp.temporal_rebin(wrf_dataset, temporal_resolution='annual')
print("KNMI_Dataset.values shape: %s" % (knmi_dataset.values.shape,))
print("WRF_Dataset.values shape: %s \n\n" % (wrf_dataset.values.shape,))
# Step 3: Spatially Regrid the Dataset Objects to a 1 degree grid.
# The spatial_boundaries() function returns the spatial extent of the dataset
print("The KNMI_Dataset spatial bounds (min_lat, max_lat, min_lon, max_lon) are: \n"
"%s\n" % (knmi_dataset.spatial_boundaries(), ))
print("The KNMI_Dataset spatial resolution (lat_resolution, lon_resolution) is: \n"
"%s\n\n" % (knmi_dataset.spatial_resolution(), ))
min_lat, max_lat, min_lon, max_lon = knmi_dataset.spatial_boundaries()
# Using the bounds we will create a new set of lats and lons on 1 degree step
new_lons = np.arange(min_lon, max_lon, 1)
# day of the month. Let's fix that real quick.
##########################################################################
knmi_dataset = dsp.normalize_dataset_datetimes(knmi_dataset, 'monthly')
wrf_dataset = dsp.normalize_dataset_datetimes(wrf_dataset, 'monthly')
# We're only going to run this evaluation over a years worth of data. We'll
# make a Bounds object and use it to subset our datasets.
##########################################################################
subset = Bounds(lat_min=-45, lat_max=42, lon_min=-24, lon_max=60,
start=datetime.datetime(1989, 1, 1), end=datetime.datetime(1989, 12, 1))
knmi_dataset = dsp.subset(knmi_dataset, subset)
wrf_dataset = dsp.subset(wrf_dataset, subset)
# Temporally re-bin the data into a monthly timestep.
##########################################################################
knmi_dataset = dsp.temporal_rebin(knmi_dataset, temporal_resolution='monthly')
wrf_dataset = dsp.temporal_rebin(wrf_dataset, temporal_resolution='monthly')
# Spatially regrid the datasets onto a 1 degree grid.
##########################################################################
# Get the bounds of the reference dataset and use it to create a new
# set of lat/lon values on a 1 degree step
# Using the bounds we will create a new set of lats and lons on 1 degree step
min_lat, max_lat, min_lon, max_lon = knmi_dataset.spatial_boundaries()
new_lons = numpy.arange(min_lon, max_lon, 1)
new_lats = numpy.arange(min_lat, max_lat, 1)
# Spatially regrid datasets using the new_lats, new_lons numpy arrays
knmi_dataset = dsp.spatial_regrid(knmi_dataset, new_lats, new_lons)
wrf_dataset = dsp.spatial_regrid(wrf_dataset, new_lats, new_lons)
""" Step 3: Resample Datasets so they are the same shape """
print("CRU31_Dataset.values shape: (times, lats, lons) - %s" % (cru31_dataset.values.shape,))
print("KNMI_Dataset.values shape: (times, lats, lons) - %s" % (knmi_dataset.values.shape,))
print("Our two datasets have a mis-match in time. We will subset on time to %s years\n" % YEARS)
# Create a Bounds object to use for subsetting
new_bounds = Bounds(min_lat, max_lat, min_lon, max_lon, start_time, end_time)
knmi_dataset = dsp.subset(new_bounds, knmi_dataset)
print("CRU31_Dataset.values shape: (times, lats, lons) - %s" % (cru31_dataset.values.shape,))
print("KNMI_Dataset.values shape: (times, lats, lons) - %s \n" % (knmi_dataset.values.shape,))
print("Temporally Rebinning the Datasets to a Single Timestep")
# To run FULL temporal Rebinning use a timedelta > 366 days. I used 999 in this example
knmi_dataset = dsp.temporal_rebin(knmi_dataset, datetime.timedelta(days=999))
cru31_dataset = dsp.temporal_rebin(cru31_dataset, datetime.timedelta(days=999))
print("KNMI_Dataset.values shape: %s" % (knmi_dataset.values.shape,))
print("CRU31_Dataset.values shape: %s \n\n" % (cru31_dataset.values.shape,))
""" Spatially Regrid the Dataset Objects to a 1/2 degree grid """
# Using the bounds we will create a new set of lats and lons on 1 degree step
new_lons = np.arange(min_lon, max_lon, 0.5)
new_lats = np.arange(min_lat, max_lat, 0.5)
# Spatially regrid datasets using the new_lats, new_lons numpy arrays
print("Spatially Regridding the KNMI_Dataset...")
knmi_dataset = dsp.spatial_regrid(knmi_dataset, new_lats, new_lons)
print("Spatially Regridding the CRU31_Dataset...")
cru31_dataset = dsp.spatial_regrid(cru31_dataset, new_lats, new_lons)
""" Step 3: Resample Datasets so they are the same shape """
print("CRU31_Dataset.values shape: (times, lats, lons) - %s" % (cru31_dataset.values.shape,))
print("KNMI_Dataset.values shape: (times, lats, lons) - %s" % (knmi_dataset.values.shape,))
print("Our two datasets have a mis-match in time. We will subset on time to %s years\n" % YEARS)
# Create a Bounds object to use for subsetting
new_bounds = Bounds(min_lat, max_lat, min_lon, max_lon, start_time, end_time)
knmi_dataset = dsp.subset(knmi_dataset, new_bounds)
print("CRU31_Dataset.values shape: (times, lats, lons) - %s" % (cru31_dataset.values.shape,))
print("KNMI_Dataset.values shape: (times, lats, lons) - %s \n" % (knmi_dataset.values.shape,))
print("Temporally Rebinning the Datasets to a Single Timestep")
# To run FULL temporal Rebinning
knmi_dataset = dsp.temporal_rebin(knmi_dataset, temporal_resolution = 'full')
cru31_dataset = dsp.temporal_rebin(cru31_dataset, temporal_resolution = 'full')
print("KNMI_Dataset.values shape: %s" % (knmi_dataset.values.shape,))
print("CRU31_Dataset.values shape: %s \n\n" % (cru31_dataset.values.shape,))
""" Spatially Regrid the Dataset Objects to a 1/2 degree grid """
# Using the bounds we will create a new set of lats and lons on 0.5 degree step
new_lons = np.arange(min_lon, max_lon, 0.5)
new_lats = np.arange(min_lat, max_lat, 0.5)
# Spatially regrid datasets using the new_lats, new_lons numpy arrays
print("Spatially Regridding the KNMI_Dataset...")
knmi_dataset = dsp.spatial_regrid(knmi_dataset, new_lats, new_lons)
print("Spatially Regridding the CRU31_Dataset...")
cru31_dataset = dsp.spatial_regrid(cru31_dataset, new_lats, new_lons)
print('OCW Benchmarks')
ocw_results = {}
with time_block(ocw_results, 'Dataset Loading'):
datasets = local.load_multiple_files(paths, 'clt')
with time_block(ocw_results, 'Domain Subsetting'):
for i, ds in enumerate(datasets):
datasets[i] = dsp.subset(ds, bnds)
with time_block(ocw_results, 'Seasonal Subsetting'):
for i, ds in enumerate(datasets):
datasets[i] = dsp.temporal_subset(ds, 9, 11)
with time_block(ocw_results, 'Resampling'):
for i, ds in enumerate(datasets):
datasets[i] = dsp.temporal_rebin(ds, 'annual')
with time_block(ocw_results, 'Regridding'):
for i, ds in enumerate(datasets):
datasets[i] = dsp.spatial_regrid(ds, new_lats, new_lons)
print(f'OCW Results: {ocw_results}')
# Plot results
matplotlib.style.use('ggplot')
df = pd.DataFrame({'OCW': ocw_results, 'BCDP': bcdp_results})
df.plot.bar(logy=True, rot=12)
for p in ax.patches:
val = np.round(p.get_height(), decimals=2)
ax.annotate(str(val), (p.get_x() + .02, p.get_height()), size=9.5)
if not 'boundary_type' in space_info:
bounds = Bounds(lat_min=min_lat,
lat_max=max_lat,
lon_min=min_lon,
lon_max=max_lon,
start=start_time,
end=end_time)
else:
bounds = Bounds(boundary_type=space_info['boundary_type'],
start=start_time,
end=end_time)
for i, dataset in enumerate(datasets):
datasets[i] = dsp.subset(dataset, bounds)
if dataset.temporal_resolution() != temporal_resolution:
datasets[i] = dsp.temporal_rebin(dataset, temporal_resolution)
# Temporally subset both observation and model datasets
# for the user specified season
month_start = time_info['month_start']
month_end = time_info['month_end']
average_each_year = time_info['average_each_year']
# For now we will treat the first listed dataset as the reference dataset for
# evaluation purposes.
for i, dataset in enumerate(datasets):
datasets[i] = dsp.temporal_subset(dataset, month_start, month_end,
average_each_year)
reference_dataset = datasets[0]
target_datasets = datasets[1:]
SEASON_MONTH_END = 12 EVAL_BOUNDS = Bounds(LAT_MIN, LAT_MAX, LON_MIN, LON_MAX, START, END) # Normalize the time values of our datasets so they fall on expected days # of the month. For example, monthly data will be normalized so that: # 15 Jan 2014, 15 Feb 2014 => 1 Jan 2014, 1 Feb 2014 ref_dataset = dsp.normalize_dataset_datetimes(ref_dataset, "monthly") target_dataset = dsp.normalize_dataset_datetimes(target_dataset, "monthly") # Subset down the evaluation datasets to our selected evaluation bounds. target_dataset = dsp.subset(EVAL_BOUNDS, target_dataset) ref_dataset = dsp.subset(EVAL_BOUNDS, ref_dataset) # Do a monthly temporal rebin of the evaluation datasets. target_dataset = dsp.temporal_rebin(target_dataset, datetime.timedelta(days=30)) ref_dataset = dsp.temporal_rebin(ref_dataset, datetime.timedelta(days=30)) # Spatially regrid onto a 1 degree lat/lon grid within our evaluation bounds. new_lats = np.arange(LAT_MIN, LAT_MAX, 1.0) new_lons = np.arange(LON_MIN, LON_MAX, 1.0) target_dataset = dsp.spatial_regrid(target_dataset, new_lats, new_lons) ref_dataset = dsp.spatial_regrid(ref_dataset, new_lats, new_lons) # Load the datasets for the evaluation. mean_bias = metrics.MeanBias() # These versions of the metrics require seasonal bounds prior to running # the metrics. You should set these values above in the evaluation # configuration section. spatial_std_dev_ratio = metrics.SeasonalSpatialStdDevRatio(month_start=SEASON_MONTH_START, month_end=SEASON_MONTH_END) pattern_correlation = metrics.SeasonalPatternCorrelation(month_start=SEASON_MONTH_START, month_end=SEASON_MONTH_END)
""" Step 1: Load Local NetCDF File into OCW Dataset Objects and store in list"""
target_datasets.append(local.load_file(FILE_1, varName, name="KNMI"))
target_datasets.append(local.load_file(FILE_2, varName, name="UC"))
target_datasets.append(local.load_file(FILE_3, varName, name="UCT"))
""" Step 2: Fetch an OCW Dataset Object from the data_source.rcmed module """
print("Working with the rcmed interface to get CRU3.1 Daily Precipitation")
# the dataset_id and the parameter id were determined from
# https://rcmes.jpl.nasa.gov/content/data-rcmes-database
CRU31 = rcmed.parameter_dataset(10, 37, LAT_MIN, LAT_MAX, LON_MIN, LON_MAX, START, END)
""" Step 3: Resample Datasets so they are the same shape """
print("Resampling datasets")
CRU31 = dsp.water_flux_unit_conversion(CRU31)
CRU31 = dsp.temporal_rebin(CRU31, datetime.timedelta(days=30))
for member, each_target_dataset in enumerate(target_datasets):
target_datasets[member] = dsp.subset(EVAL_BOUNDS, target_datasets[member])
target_datasets[member] = dsp.water_flux_unit_conversion(target_datasets[member])
target_datasets[member] = dsp.temporal_rebin(target_datasets[member], datetime.timedelta(days=30))
""" Spatially Regrid the Dataset Objects to a user defined grid """
# Using the bounds we will create a new set of lats and lons
print("Regridding datasets")
new_lats = np.arange(LAT_MIN, LAT_MAX, gridLatStep)
new_lons = np.arange(LON_MIN, LON_MAX, gridLonStep)
CRU31 = dsp.spatial_regrid(CRU31, new_lats, new_lons)
for member, each_target_dataset in enumerate(target_datasets):
print 'Maximum overlap period'
print 'start_time:', start_time
print 'end_time:', end_time
if temporal_resolution == 'monthly' and end_time.day !=1:
end_time = end_time.replace(day=1)
if ref_data_info['data_source'] == 'rcmed':
min_lat = np.max([min_lat, ref_dataset.lats.min()])
max_lat = np.min([max_lat, ref_dataset.lats.max()])
min_lon = np.max([min_lon, ref_dataset.lons.min()])
max_lon = np.min([max_lon, ref_dataset.lons.max()])
bounds = Bounds(min_lat, max_lat, min_lon, max_lon, start_time, end_time)
ref_dataset = dsp.subset(bounds,ref_dataset)
if ref_dataset.temporal_resolution() != temporal_resolution:
ref_dataset = dsp.temporal_rebin(ref_dataset, temporal_resolution)
for idata,dataset in enumerate(model_datasets):
model_datasets[idata] = dsp.subset(bounds,dataset)
if dataset.temporal_resolution() != temporal_resolution:
model_datasets[idata] = dsp.temporal_rebin(dataset, temporal_resolution)
# Temporaly subset both observation and model datasets for the user specified season
month_start = time_info['month_start']
month_end = time_info['month_end']
average_each_year = time_info['average_each_year']
ref_dataset = dsp.temporal_subset(month_start, month_end,ref_dataset,average_each_year)
for idata,dataset in enumerate(model_datasets):
model_datasets[idata] = dsp.temporal_subset(month_start, month_end,dataset,average_each_year)
# generate grid points for regridding
def test_invalid_unit_rebin(self):
with self.assertRaises(ValueError):
dp.temporal_rebin(self.two_years_daily_dataset, "days")
print(
"Our two datasets have a mis-match in time. We will subset on time to %s years\n"
% YEARS)
# Create a Bounds object to use for subsetting
new_bounds = Bounds(min_lat, max_lat, min_lon, max_lon, start_time, end_time)
knmi_dataset = dsp.subset(new_bounds, knmi_dataset)
print("CRU31_Dataset.values shape: (times, lats, lons) - %s" %
(cru31_dataset.values.shape, ))
print("KNMI_Dataset.values shape: (times, lats, lons) - %s \n" %
(knmi_dataset.values.shape, ))
print("Temporally Rebinning the Datasets to a Single Timestep")
# To run FULL temporal Rebinning use a timedelta > 366 days. I used 999 in this example
knmi_dataset = dsp.temporal_rebin(knmi_dataset, datetime.timedelta(days=999))
cru31_dataset = dsp.temporal_rebin(cru31_dataset, datetime.timedelta(days=999))
print("KNMI_Dataset.values shape: %s" % (knmi_dataset.values.shape, ))
print("CRU31_Dataset.values shape: %s \n\n" % (cru31_dataset.values.shape, ))
""" Spatially Regrid the Dataset Objects to a 1/2 degree grid """
# Using the bounds we will create a new set of lats and lons on 1 degree step
new_lons = np.arange(min_lon, max_lon, 0.5)
new_lats = np.arange(min_lat, max_lat, 0.5)
# Spatially regrid datasets using the new_lats, new_lons numpy arrays
print("Spatially Regridding the KNMI_Dataset...")
knmi_dataset = dsp.spatial_regrid(knmi_dataset, new_lats, new_lons)
print("Spatially Regridding the CRU31_Dataset...")
cru31_dataset = dsp.spatial_regrid(cru31_dataset, new_lats, new_lons)
print("Final shape of the KNMI_Dataset:%s" % (knmi_dataset.values.shape, ))
if not 'boundary_type' in space_info:
bounds = Bounds(lat_min=min_lat,
lat_max=max_lat,
lon_min=min_lon,
lon_max=max_lon,
start=start_time,
end=end_time)
else:
bounds = Bounds(boundary_type=space_info['boundary_type'],
start=start_time,
end=end_time)
for i, dataset in enumerate(obs_datasets):
obs_datasets[i] = dsp.subset(dataset, bounds)
if dataset.temporal_resolution() != temporal_resolution:
obs_datasets[i] = dsp.temporal_rebin(dataset, temporal_resolution)
for i, dataset in enumerate(model_datasets):
model_datasets[i] = dsp.subset(dataset, bounds)
if dataset.temporal_resolution() != temporal_resolution:
model_datasets[i] = dsp.temporal_rebin(dataset, temporal_resolution)
# Temporally subset both observation and model datasets
# for the user specified season
month_start = time_info['month_start']
month_end = time_info['month_end']
average_each_year = time_info['average_each_year']
# TODO: Fully support multiple observation / reference datasets.
# For now we will only use the first reference dataset listed in the config file
obs_dataset = obs_datasets[0]
if not 'boundary_type' in space_info:
bounds = Bounds(lat_min=min_lat,
lat_max=max_lat,
lon_min=min_lon,
lon_max=max_lon,
start=start_time,
end=end_time)
else:
bounds = Bounds(boundary_type=space_info['boundary_type'],
start=start_time,
end=end_time)
for i, dataset in enumerate(datasets):
datasets[i] = dsp.subset(dataset, bounds)
if dataset.temporal_resolution() != temporal_resolution:
datasets[i] = dsp.temporal_rebin(datasets[i], temporal_resolution)
# Temporally subset both observation and model datasets
# for the user specified season
month_start = time_info['month_start']
month_end = time_info['month_end']
average_each_year = time_info['average_each_year']
# For now we will treat the first listed dataset as the reference dataset for
# evaluation purposes.
for i, dataset in enumerate(datasets):
datasets[i] = dsp.temporal_subset(dataset, month_start, month_end,
average_each_year)
reference_dataset = datasets[0]
target_datasets = datasets[1:]
def run_evaluation():
''' Run an OCW Evaluation.
*run_evaluation* expects the Evaluation parameters to be POSTed in
the following format.
.. sourcecode:: javascript
{
reference_dataset: {
// Id that tells us how we need to load this dataset.
'data_source_id': 1 == local, 2 == rcmed,
// Dict of data_source specific identifying information.
//
// if data_source_id == 1 == local:
// {
// 'id': The path to the local file on the server for loading.
// 'var_name': The variable data to pull from the file.
// 'lat_name': The latitude variable name.
// 'lon_name': The longitude variable name.
// 'time_name': The time variable name
// 'name': Optional dataset name
// }
//
// if data_source_id == 2 == rcmed:
// {
// 'dataset_id': The dataset id to grab from RCMED.
// 'parameter_id': The variable id value used by RCMED.
// 'name': Optional dataset name
// }
'dataset_info': {..}
},
// The list of target datasets to use in the Evaluation. The data
// format for the dataset objects should be the same as the
// reference_dataset above.
'target_datasets': [{...}, {...}, ...],
// All the datasets are re-binned to the reference dataset
// before being added to an experiment. This step (in degrees)
// is used when re-binning both the reference and target datasets.
'spatial_rebin_lat_step': The lat degree step. Integer > 0,
// Same as above, but for lon
'spatial_rebin_lon_step': The lon degree step. Integer > 0,
// The temporal resolution to use when doing a temporal re-bin
// This is a timedelta of days to use so daily == 1, monthly is
// (1, 31], annual/yearly is (31, 366], and full is anything > 366.
'temporal_resolution': Integer in range(1, 999),
// A list of the metric class names to use in the evaluation. The
// names must match the class name exactly.
'metrics': [Bias, TemporalStdDev, ...]
// The bounding values used in the Evaluation. Note that lat values
// should range from -180 to 180 and lon values from -90 to 90.
'start_time': start time value in the format '%Y-%m-%d %H:%M:%S',
'end_time': end time value in the format '%Y-%m-%d %H:%M:%S',
'lat_min': The minimum latitude value,
'lat_max': The maximum latitude value,
'lon_min': The minimum longitude value,
'lon_max': The maximum longitude value,
// NOTE: At the moment, subregion support is fairly minimal. This
// will be addressed in the future. Ideally, the user should be able
// to load a file that they have locally. That would change the
// format that this data is passed.
'subregion_information': Path to a subregion file on the server.
}
'''
# TODO: validate input parameters and return an error if not valid
eval_time_stamp = datetime.now().strftime('%Y-%m-%d_%H-%M-%S')
data = request.json
eval_bounds = {
'start_time': datetime.strptime(data['start_time'], '%Y-%m-%d %H:%M:%S'),
'end_time': datetime.strptime(data['end_time'], '%Y-%m-%d %H:%M:%S'),
'lat_min': float(data['lat_min']),
'lat_max': float(data['lat_max']),
'lon_min': float(data['lon_min']),
'lon_max': float(data['lon_max'])
}
# Load all the datasets
ref_dataset = _process_dataset_object(data['reference_dataset'], eval_bounds)
target_datasets = [_process_dataset_object(obj, eval_bounds)
for obj
in data['target_datasets']]
# Normalize the dataset time values so they break on consistent days of the
# month or time of the day, depending on how they will be rebinned.
resolution = data['temporal_resolution']
time_delta = timedelta(days=resolution)
time_step = 'daily' if resolution == 1 else 'monthly'
ref_dataset = dsp.normalize_dataset_datetimes(ref_dataset, time_step)
target_datasets = [dsp.normalize_dataset_datetimes(ds, time_step)
for ds in target_datasets]
# Subset the datasets
start = eval_bounds['start_time']
end = eval_bounds['end_time']
# Normalize all the values to the first of the month if we're not
# dealing with daily data. This will ensure that a valid subregion
# isn't considered out of bounds do to a dataset's time values
# being shifted to the first of the month.
if time_step != 'daily':
if start.day != 1:
day_offset = start.day - 1
start -= timedelta(days=day_offset)
if end.day != 1:
day_offset = end.day - 1
end -= timedelta(days=day_offset)
subset = Bounds(eval_bounds['lat_min'],
eval_bounds['lat_max'],
eval_bounds['lon_min'],
eval_bounds['lon_max'],
start,
end)
ref_dataset = dsp.safe_subset(subset, ref_dataset)
target_datasets = [dsp.safe_subset(subset, ds)
for ds
in target_datasets]
# Do temporal re-bin based off of passed resolution
ref_dataset = dsp.temporal_rebin(ref_dataset, time_delta)
target_datasets = [dsp.temporal_rebin(ds, time_delta)
for ds
in target_datasets]
# Do spatial re=bin based off of reference dataset + lat/lon steps
lat_step = data['spatial_rebin_lat_step']
lon_step = data['spatial_rebin_lon_step']
lat_bins, lon_bins = _calculate_new_latlon_bins(eval_bounds,
lat_step,
lon_step)
ref_dataset = dsp.spatial_regrid(ref_dataset, lat_bins, lon_bins)
target_datasets = [dsp.spatial_regrid(ds, lat_bins, lon_bins)
for ds
in target_datasets]
# Load metrics
loaded_metrics = _load_metrics(data['metrics'])
# Prime evaluation object with data
evaluation = Evaluation(ref_dataset, target_datasets, loaded_metrics)
# Run evaluation
evaluation.run()
# Plot
_generate_evaluation_plots(evaluation, lat_bins, lon_bins, eval_time_stamp)
return json.dumps({'eval_work_dir': eval_time_stamp})
def test_monthly_to_annual_rebin(self):
annual_dataset = dp.temporal_rebin(
self.ten_year_monthly_dataset, "annual")
np.testing.assert_array_equal(
annual_dataset.times, self.ten_year_annual_times)
def test_monthly_to_full_rebin(self):
full_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset,
datetime.timedelta(days=3650))
full_times = [datetime.datetime(2004, 12, 16)]
self.assertEqual(full_dataset.times, full_times)
def test_monthly_to_full_rebin(self):
full_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset, "full")
full_times = [datetime.datetime(2005, 1, 1)]
self.assertEqual(full_dataset.times, full_times)
target_datasets.append(local.load_file(FILE_2, varName, name="REGM3"))
target_datasets.append(local.load_file(FILE_3, varName, name="UCT"))
""" Step 2: Fetch an OCW Dataset Object from the data_source.rcmed module """
print(
"Working with the rcmed interface to get CRU3.1 Monthly Mean Precipitation"
)
# the dataset_id and the parameter id were determined from
# https://rcmes.jpl.nasa.gov/content/data-rcmes-database
CRU31 = rcmed.parameter_dataset(10, 37, LAT_MIN, LAT_MAX, LON_MIN, LON_MAX,
START, END)
""" Step 3: Resample Datasets so they are the same shape """
print("Resampling datasets ...")
print("... on units")
CRU31 = dsp.water_flux_unit_conversion(CRU31)
print("... temporal")
CRU31 = dsp.temporal_rebin(CRU31, temporal_resolution='monthly')
for member, each_target_dataset in enumerate(target_datasets):
target_datasets[member] = dsp.water_flux_unit_conversion(
target_datasets[member])
target_datasets[member] = dsp.temporal_rebin(target_datasets[member],
temporal_resolution='monthly')
target_datasets[member] = dsp.subset(EVAL_BOUNDS, target_datasets[member])
#Regrid
print("... regrid")
new_lats = np.arange(LAT_MIN, LAT_MAX, gridLatStep)
new_lons = np.arange(LON_MIN, LON_MAX, gridLonStep)
CRU31 = dsp.spatial_regrid(CRU31, new_lats, new_lons)
for member, each_target_dataset in enumerate(target_datasets):
def run_screen(model_datasets, models_info, observations_info,
overlap_start_time, overlap_end_time, overlap_min_lat,
overlap_max_lat, overlap_min_lon, overlap_max_lon,
temp_grid_setting, spatial_grid_setting, working_directory, plot_title):
'''Generates screen to show running evaluation process.
:param model_datasets: list of model dataset objects
:type model_datasets: list
:param models_info: list of dictionaries that contain information for each model
:type models_info: list
:param observations_info: list of dictionaries that contain information for each observation
:type observations_info: list
:param overlap_start_time: overlap start time between model and obs start time
:type overlap_start_time: datetime
:param overlap_end_time: overlap end time between model and obs end time
:type overlap_end_time: float
:param overlap_min_lat: overlap minimum lat between model and obs minimum lat
:type overlap_min_lat: float
:param overlap_max_lat: overlap maximum lat between model and obs maximum lat
:type overlap_max_lat: float
:param overlap_min_lon: overlap minimum lon between model and obs minimum lon
:type overlap_min_lon: float
:param overlap_max_lon: overlap maximum lon between model and obs maximum lon
:type overlap_max_lon: float
:param temp_grid_setting: temporal grid option such as hourly, daily, monthly and annually
:type temp_grid_setting: string
:param spatial_grid_setting:
:type spatial_grid_setting: string
:param working_directory: path to a directory for storring outputs
:type working_directory: string
:param plot_title: Title for plot
:type plot_title: string
'''
option = None
if option != "0":
ready_screen("manage_obs_screen")
y = screen.getmaxyx()[0]
screen.addstr(2, 2, "Evaluation started....")
screen.refresh()
OUTPUT_PLOT = "plot"
dataset_id = int(observations_info[0]['dataset_id']) #just accepts one dataset at this time
parameter_id = int(observations_info[0]['parameter_id']) #just accepts one dataset at this time
new_bounds = Bounds(overlap_min_lat, overlap_max_lat, overlap_min_lon, overlap_max_lon, overlap_start_time, overlap_end_time)
model_dataset = dsp.subset(new_bounds, model_datasets[0]) #just accepts one model at this time
#Getting bound info of subseted model file to retrive obs data with same bound as subseted model
new_model_spatial_bounds = model_dataset.spatial_boundaries()
new_model_temp_bounds = model_dataset.time_range()
new_min_lat = new_model_spatial_bounds[0]
new_max_lat = new_model_spatial_bounds[1]
new_min_lon = new_model_spatial_bounds[2]
new_max_lon = new_model_spatial_bounds[3]
new_start_time = new_model_temp_bounds[0]
new_end_time = new_model_temp_bounds[1]
screen.addstr(4, 4, "Retrieving data...")
screen.refresh()
#Retrieve obs data
obs_dataset = rcmed.parameter_dataset(
dataset_id,
parameter_id,
new_min_lat,
new_max_lat,
new_min_lon,
new_max_lon,
new_start_time,
new_end_time)
screen.addstr(4, 4, "--> Data retrieved.")
screen.refresh()
screen.addstr(5, 4, "Temporally regridding...")
screen.refresh()
if temp_grid_setting.lower() == 'hourly':
days = 0.5
elif temp_grid_setting.lower() == 'daily':
days = 1
elif temp_grid_setting.lower() == 'monthly':
days = 31
else:
days = 365
model_dataset = dsp.temporal_rebin(model_dataset, timedelta(days))
obs_dataset = dsp.temporal_rebin(obs_dataset, timedelta(days))
screen.addstr(5, 4, "--> Temporally regridded.")
screen.refresh()
new_lats = np.arange(new_min_lat, new_max_lat, spatial_grid_setting)
new_lons = np.arange(new_min_lon, new_max_lon, spatial_grid_setting)
screen.addstr(6, 4, "Spatially regridding...")
screen.refresh()
spatial_gridded_model = dsp.spatial_regrid(model_dataset, new_lats, new_lons)
spatial_gridded_obs = dsp.spatial_regrid(obs_dataset, new_lats, new_lons)
screen.addstr(6, 4, "--> Spatially regridded.")
screen.refresh()
screen.addstr(7, 4, "Setting up metrics...")
screen.refresh()
bias = metrics.Bias()
bias_evaluation = evaluation.Evaluation(spatial_gridded_model, [spatial_gridded_obs], [bias])
screen.addstr(7, 4, "--> Metrics setting done.")
screen.refresh()
screen.addstr(8, 4, "Running evaluation.....")
screen.refresh()
bias_evaluation.run()
results = bias_evaluation.results[0][0]
screen.addstr(8, 4, "--> Evaluation Finished.")
screen.refresh()
screen.addstr(9, 4, "Generating plots....")
screen.refresh()
lats = new_lats
lons = new_lons
gridshape = (1, 1)
sub_titles = [""] #No subtitle set for now
if not os.path.exists(working_directory):
os.makedirs(working_directory)
for i in range(len(results)):
fname = working_directory + OUTPUT_PLOT + str(i)
plotter.draw_contour_map(results[i], lats, lons, fname,
gridshape=gridshape, ptitle=plot_title,
subtitles=sub_titles)
screen.addstr(9, 4, "--> Plots generated.")
screen.refresh()
screen.addstr(y-2, 1, "Press 'enter' to Exit: ")
option = screen.getstr()
print 'Maximum overlap period'
print 'start_time:', start_time
print 'end_time:', end_time
if temporal_resolution == 'monthly' and end_time.day !=1:
end_time = end_time.replace(day=1)
if ref_data_info['data_source'] == 'rcmed':
min_lat = np.max([min_lat, ref_dataset.lats.min()])
max_lat = np.min([max_lat, ref_dataset.lats.max()])
min_lon = np.max([min_lon, ref_dataset.lons.min()])
max_lon = np.min([max_lon, ref_dataset.lons.max()])
bounds = Bounds(min_lat, max_lat, min_lon, max_lon, start_time, end_time)
ref_dataset = dsp.subset(bounds,ref_dataset)
if ref_dataset.temporal_resolution() != temporal_resolution:
ref_dataset = dsp.temporal_rebin(ref_dataset, temporal_resolution)
for idata,dataset in enumerate(model_datasets):
model_datasets[idata] = dsp.subset(bounds,dataset)
if dataset.temporal_resolution() != temporal_resolution:
model_datasets[idata] = dsp.temporal_rebin(dataset, temporal_resolution)
# Temporaly subset both observation and model datasets for the user specified season
month_start = time_info['month_start']
month_end = time_info['month_end']
average_each_year = time_info['average_each_year']
ref_dataset = dsp.temporal_subset(month_start, month_end,ref_dataset,average_each_year)
for idata,dataset in enumerate(model_datasets):
model_datasets[idata] = dsp.temporal_subset(month_start, month_end,dataset,average_each_year)
# generate grid points for regridding
def test_monthly_to_full_rebin(self):
full_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset, "full")
full_times = [datetime.datetime(2005, 1, 1)]
self.assertEqual(full_dataset.times, full_times)
print("Fetching data from RCMED...")
cru31_dataset = rcmed.parameter_dataset(dataset_id,
parameter_id,
min_lat,
max_lat,
min_lon,
max_lon,
start_time,
end_time)
""" Step 3: Resample Datasets so they are the same shape """
print("Temporally Rebinning the Datasets to an Annual Timestep")
# To run annual temporal Rebinning use a timedelta of 360 days.
knmi_dataset = dsp.temporal_rebin(knmi_dataset, datetime.timedelta(days=360))
wrf311_dataset = dsp.temporal_rebin(wrf311_dataset, datetime.timedelta(days=360))
cru31_dataset = dsp.temporal_rebin(cru31_dataset, datetime.timedelta(days=360))
# Running Temporal Rebin early helps negate the issue of datasets being on different
# days of the month (1st vs. 15th)
# Create a Bounds object to use for subsetting
new_bounds = Bounds(min_lat, max_lat, min_lon, max_lon, start_time, end_time)
# Subset our model datasets so they are the same size
knmi_dataset = dsp.subset(new_bounds, knmi_dataset)
wrf311_dataset = dsp.subset(new_bounds, wrf311_dataset)
""" Spatially Regrid the Dataset Objects to a 1/2 degree grid """
# Using the bounds we will create a new set of lats and lons on 1/2 degree step
new_lons = np.arange(min_lon, max_lon, 0.5)
# Set the Time Range to be the year 1989
start_time = datetime.datetime(1989, 1, 1)
end_time = datetime.datetime(1989, 12, 1)
print("Time Range is: %s to %s" %
(start_time.strftime("%Y-%m-%d"), end_time.strftime("%Y-%m-%d")))
print("Fetching data from RCMED...")
cru31_dataset = rcmed.parameter_dataset(dataset_id, parameter_id, min_lat,
max_lat, min_lon, max_lon, start_time,
end_time)
""" Step 3: Resample Datasets so they are the same shape """
print("Temporally Rebinning the Datasets to an Annual Timestep")
# To run annual temporal Rebinning use a timedelta of 360 days.
knmi_dataset = dsp.temporal_rebin(knmi_dataset, datetime.timedelta(days=360))
wrf311_dataset = dsp.temporal_rebin(wrf311_dataset,
datetime.timedelta(days=360))
cru31_dataset = dsp.temporal_rebin(cru31_dataset, datetime.timedelta(days=360))
# Running Temporal Rebin early helps negate the issue of datasets being on different
# days of the month (1st vs. 15th)
# Create a Bounds object to use for subsetting
new_bounds = Bounds(min_lat, max_lat, min_lon, max_lon, start_time, end_time)
# Subset our model datasets so they are the same size
knmi_dataset = dsp.subset(new_bounds, knmi_dataset)
wrf311_dataset = dsp.subset(new_bounds, wrf311_dataset)
""" Spatially Regrid the Dataset Objects to a 1/2 degree grid """
# Using the bounds we will create a new set of lats and lons on 1/2 degree step
new_lons = np.arange(min_lon, max_lon, 0.5)
def test_monthly_to_annual_rebin(self):
annual_dataset = dp.temporal_rebin(
self.ten_year_monthly_dataset, "annual")
np.testing.assert_array_equal(
annual_dataset.times, self.ten_year_annual_times)
target_datasets.append(local.load_file(FILE_3, varName, name="UCT"))
""" Step 2: Fetch an OCW Dataset Object from the data_source.rcmed module """
print("Working with the rcmed interface to get CRU3.1 Monthly Mean Precipitation")
# the dataset_id and the parameter id were determined from
# https://rcmes.jpl.nasa.gov/content/data-rcmes-database
CRU31 = rcmed.parameter_dataset(
10, 37, LAT_MIN, LAT_MAX, LON_MIN, LON_MAX, START, END)
""" Step 3: Resample Datasets so they are the same shape """
print("Resampling datasets ...")
print("... on units")
CRU31 = dsp.water_flux_unit_conversion(CRU31)
print("... temporal")
CRU31 = dsp.temporal_rebin(CRU31, temporal_resolution='monthly')
for member, each_target_dataset in enumerate(target_datasets):
target_datasets[member] = dsp.water_flux_unit_conversion(target_datasets[
member])
target_datasets[member] = dsp.temporal_rebin(
target_datasets[member], temporal_resolution='monthly')
target_datasets[member] = dsp.subset(target_datasets[member], EVAL_BOUNDS)
# Regrid
print("... regrid")
new_lats = np.arange(LAT_MIN, LAT_MAX, gridLatStep)
new_lons = np.arange(LON_MIN, LON_MAX, gridLonStep)
CRU31 = dsp.spatial_regrid(CRU31, new_lats, new_lons)
for member, each_target_dataset in enumerate(target_datasets):
def test_non_rebin(self):
"""This will take a monthly dataset and ask for a monthly rebin of 28 days. The resulting
dataset should have the same time values"""
monthly_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset, datetime.timedelta(days=28))
good_times = self.ten_year_monthly_dataset.times
np.testing.assert_array_equal(monthly_dataset.times, good_times)
def test_daily_to_daily_rebin(self):
daily_dataset = dp.temporal_rebin(
self.two_years_daily_dataset, "daily")
np.testing.assert_array_equal(
daily_dataset.times, self.two_years_daily_dataset.times)
def test_daily_to_daily_rebin(self):
daily_dataset = dp.temporal_rebin(
self.two_years_daily_dataset, "daily")
np.testing.assert_array_equal(
daily_dataset.times, self.two_years_daily_dataset.times)
def test_invalid_unit_rebin(self):
with self.assertRaises(ValueError):
dp.temporal_rebin(self.two_years_daily_dataset, "days")
def run_evaluation():
''' Run an OCW Evaluation.
*run_evaluation* expects the Evaluation parameters to be POSTed in
the following format.
.. sourcecode:: javascript
{
reference_dataset: {
// Id that tells us how we need to load this dataset.
'data_source_id': 1 == local, 2 == rcmed,
// Dict of data_source specific identifying information.
//
// if data_source_id == 1 == local:
// {
// 'id': The path to the local file on the server for loading.
// 'var_name': The variable data to pull from the file.
// 'lat_name': The latitude variable name.
// 'lon_name': The longitude variable name.
// 'time_name': The time variable name
// 'name': Optional dataset name
// }
//
// if data_source_id == 2 == rcmed:
// {
// 'dataset_id': The dataset id to grab from RCMED.
// 'parameter_id': The variable id value used by RCMED.
// 'name': Optional dataset name
// }
'dataset_info': {..}
},
// The list of target datasets to use in the Evaluation. The data
// format for the dataset objects should be the same as the
// reference_dataset above.
'target_datasets': [{...}, {...}, ...],
// All the datasets are re-binned to the reference dataset
// before being added to an experiment. This step (in degrees)
// is used when re-binning both the reference and target datasets.
'spatial_rebin_lat_step': The lat degree step. Integer > 0,
// Same as above, but for lon
'spatial_rebin_lon_step': The lon degree step. Integer > 0,
// The temporal resolution to use when doing a temporal re-bin
// This is a timedelta of days to use so daily == 1, monthly is
// (1, 31], annual/yearly is (31, 366], and full is anything > 366.
'temporal_resolution': Integer in range(1, 999),
// A list of the metric class names to use in the evaluation. The
// names must match the class name exactly.
'metrics': [Bias, TemporalStdDev, ...]
// The bounding values used in the Evaluation. Note that lat values
// should range from -180 to 180 and lon values from -90 to 90.
'start_time': start time value in the format '%Y-%m-%d %H:%M:%S',
'end_time': end time value in the format '%Y-%m-%d %H:%M:%S',
'lat_min': The minimum latitude value,
'lat_max': The maximum latitude value,
'lon_min': The minimum longitude value,
'lon_max': The maximum longitude value,
// NOTE: At the moment, subregion support is fairly minimal. This
// will be addressed in the future. Ideally, the user should be able
// to load a file that they have locally. That would change the
// format that this data is passed.
'subregion_information': Path to a subregion file on the server.
}
'''
# TODO: validate input parameters and return an error if not valid
eval_time_stamp = datetime.now().strftime('%Y-%m-%d_%H-%M-%S')
data = request.json
eval_bounds = {
'start_time': datetime.strptime(data['start_time'], '%Y-%m-%d %H:%M:%S'),
'end_time': datetime.strptime(data['end_time'], '%Y-%m-%d %H:%M:%S'),
'lat_min': float(data['lat_min']),
'lat_max': float(data['lat_max']),
'lon_min': float(data['lon_min']),
'lon_max': float(data['lon_max'])
}
# Load all the datasets
ref_dataset = _process_dataset_object(data['reference_dataset'], eval_bounds)
target_datasets = [_process_dataset_object(obj, eval_bounds)
for obj
in data['target_datasets']]
# Normalize the dataset time values so they break on consistent days of the
# month or time of the day, depending on how they will be rebinned.
resolution = data['temporal_resolution']
time_delta = timedelta(days=resolution)
time_step = 'daily' if resolution == 1 else 'monthly'
ref_dataset = dsp.normalize_dataset_datetimes(ref_dataset, time_step)
target_datasets = [dsp.normalize_dataset_datetimes(ds, time_step)
for ds in target_datasets]
# Subset the datasets
start = eval_bounds['start_time']
end = eval_bounds['end_time']
# Normalize all the values to the first of the month if we're not
# dealing with daily data. This will ensure that a valid subregion
# isn't considered out of bounds do to a dataset's time values
# being shifted to the first of the month.
if time_step != 'daily':
if start.day != 1:
day_offset = start.day - 1
start -= timedelta(days=day_offset)
if end.day != 1:
day_offset = end.day - 1
end -= timedelta(days=day_offset)
subset = Bounds(eval_bounds['lat_min'],
eval_bounds['lat_max'],
eval_bounds['lon_min'],
eval_bounds['lon_max'],
start,
end)
ref_dataset = dsp.safe_subset(ref_dataset, subset)
target_datasets = [dsp.safe_subset(ds, subset)
for ds
in target_datasets]
# Do temporal re-bin based off of passed resolution
ref_dataset = dsp.temporal_rebin(ref_dataset, time_delta)
target_datasets = [dsp.temporal_rebin(ds, time_delta)
for ds
in target_datasets]
# Do spatial re=bin based off of reference dataset + lat/lon steps
lat_step = data['spatial_rebin_lat_step']
lon_step = data['spatial_rebin_lon_step']
lat_bins, lon_bins = _calculate_new_latlon_bins(eval_bounds,
lat_step,
lon_step)
ref_dataset = dsp.spatial_regrid(ref_dataset, lat_bins, lon_bins)
target_datasets = [dsp.spatial_regrid(ds, lat_bins, lon_bins)
for ds
in target_datasets]
# Load metrics
loaded_metrics = _load_metrics(data['metrics'])
# Prime evaluation object with data
evaluation = Evaluation(ref_dataset, target_datasets, loaded_metrics)
# Run evaluation
evaluation.run()
# Plot
_generate_evaluation_plots(evaluation, lat_bins, lon_bins, eval_time_stamp)
return json.dumps({'eval_work_dir': eval_time_stamp})
def test_monthly_to_full_rebin(self):
full_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset, datetime.timedelta(days=3650))
full_times = [datetime.datetime(2004, 12, 16)]
self.assertEqual(full_dataset.times, full_times)
new_bounds = Bounds(lat_min=min_lat,
lat_max=max_lat,
lon_min=min_lon,
lon_max=max_lon,
start=start_time,
end=end_time)
knmi_dataset = dsp.subset(knmi_dataset, new_bounds)
print("CRU31_Dataset.values shape: (times, lats, lons) - %s" %
(cru31_dataset.values.shape, ))
print("KNMI_Dataset.values shape: (times, lats, lons) - %s \n" %
(knmi_dataset.values.shape, ))
print("Temporally Rebinning the Datasets to a Single Timestep")
# To run FULL temporal Rebinning
knmi_dataset = dsp.temporal_rebin(knmi_dataset, temporal_resolution='full')
cru31_dataset = dsp.temporal_rebin(cru31_dataset, temporal_resolution='full')
print("KNMI_Dataset.values shape: %s" % (knmi_dataset.values.shape, ))
print("CRU31_Dataset.values shape: %s \n\n" % (cru31_dataset.values.shape, ))
""" Spatially Regrid the Dataset Objects to a 1/2 degree grid """
# Using the bounds we will create a new set of lats and lons on 0.5 degree step
new_lons = np.arange(min_lon, max_lon, 0.5)
new_lats = np.arange(min_lat, max_lat, 0.5)
# Spatially regrid datasets using the new_lats, new_lons numpy arrays
print("Spatially Regridding the KNMI_Dataset...")
knmi_dataset = dsp.spatial_regrid(knmi_dataset, new_lats, new_lons)
print("Spatially Regridding the CRU31_Dataset...")
cru31_dataset = dsp.spatial_regrid(cru31_dataset, new_lats, new_lons)
print("Final shape of the KNMI_Dataset:%s" % (knmi_dataset.values.shape, ))
def test_monthly_to_annual_rebin(self):
annual_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset, datetime.timedelta(days=365))
np.testing.assert_array_equal(annual_dataset.times, self.ten_year_annual_times)
print("Fetching data from RCMED...")
cru31_dataset = rcmed.parameter_dataset(dataset_id,
parameter_id,
min_lat,
max_lat,
min_lon,
max_lon,
start_time,
end_time)
""" Step 3: Resample Datasets so they are the same shape """
print("Temporally Rebinning the Datasets to an Annual Timestep")
# To run annual temporal Rebinning,
knmi_dataset = dsp.temporal_rebin(knmi_dataset, temporal_resolution = 'annual')
wrf311_dataset = dsp.temporal_rebin(wrf311_dataset, temporal_resolution = 'annual')
cru31_dataset = dsp.temporal_rebin(cru31_dataset, temporal_resolution = 'annual')
# Running Temporal Rebin early helps negate the issue of datasets being on different
# days of the month (1st vs. 15th)
# Create a Bounds object to use for subsetting
new_bounds = Bounds(min_lat, max_lat, min_lon, max_lon, start_time, end_time)
# Subset our model datasets so they are the same size
knmi_dataset = dsp.subset(knmi_dataset, new_bounds)
wrf311_dataset = dsp.subset(wrf311_dataset, new_bounds)
""" Spatially Regrid the Dataset Objects to a 1/2 degree grid """
# Using the bounds we will create a new set of lats and lons on 1/2 degree step
new_lons = np.arange(min_lon, max_lon, 0.5)
def test_monthly_to_annual_rebin(self):
annual_dataset = dp.temporal_rebin(self.ten_year_monthly_dataset,
datetime.timedelta(days=365))
np.testing.assert_array_equal(annual_dataset.times,
self.ten_year_annual_times)
# We're only going to run this evaluation over a years worth of data. We'll
# make a Bounds object and use it to subset our datasets.
##########################################################################
subset = Bounds(lat_min=-45,
lat_max=42,
lon_min=-24,
lon_max=60,
start=datetime.datetime(1989, 1, 1),
end=datetime.datetime(1989, 12, 1))
knmi_dataset = dsp.subset(knmi_dataset, subset)
wrf_dataset = dsp.subset(wrf_dataset, subset)
# Temporally re-bin the data into a monthly timestep.
##########################################################################
knmi_dataset = dsp.temporal_rebin(knmi_dataset, temporal_resolution='monthly')
wrf_dataset = dsp.temporal_rebin(wrf_dataset, temporal_resolution='monthly')
# Spatially regrid the datasets onto a 1 degree grid.
##########################################################################
# Get the bounds of the reference dataset and use it to create a new
# set of lat/lon values on a 1 degree step
# Using the bounds we will create a new set of lats and lons on 1 degree step
min_lat, max_lat, min_lon, max_lon = knmi_dataset.spatial_boundaries()
new_lons = numpy.arange(min_lon, max_lon, 1)
new_lats = numpy.arange(min_lat, max_lat, 1)
# Spatially regrid datasets using the new_lats, new_lons numpy arrays
knmi_dataset = dsp.spatial_regrid(knmi_dataset, new_lats, new_lons)
wrf_dataset = dsp.spatial_regrid(wrf_dataset, new_lats, new_lons)
target_datasets.append(local.load_file(FILE_2, varName, name="REGM3"))
target_datasets.append(local.load_file(FILE_3, varName, name="UCT"))
""" Step 2: Fetch an OCW Dataset Object from the data_source.rcmed module """
print("Working with the rcmed interface to get CRU3.1 Monthly Mean Precipitation")
# the dataset_id and the parameter id were determined from
# https://rcmes.jpl.nasa.gov/content/data-rcmes-database
CRU31 = rcmed.parameter_dataset(10, 37, LAT_MIN, LAT_MAX, LON_MIN, LON_MAX, START, END)
""" Step 3: Resample Datasets so they are the same shape """
print("Resampling datasets ...")
print("... on units")
CRU31 = dsp.water_flux_unit_conversion(CRU31)
print("... temporal")
CRU31 = dsp.temporal_rebin(CRU31, datetime.timedelta(days=30))
for member, each_target_dataset in enumerate(target_datasets):
target_datasets[member] = dsp.water_flux_unit_conversion(target_datasets[member])
target_datasets[member] = dsp.temporal_rebin(target_datasets[member], datetime.timedelta(days=30))
target_datasets[member] = dsp.subset(EVAL_BOUNDS, target_datasets[member])
#Regrid
print("... regrid")
new_lats = np.arange(LAT_MIN, LAT_MAX, gridLatStep)
new_lons = np.arange(LON_MIN, LON_MAX, gridLonStep)
CRU31 = dsp.spatial_regrid(CRU31, new_lats, new_lons)
for member, each_target_dataset in enumerate(target_datasets):
target_datasets[member] = dsp.spatial_regrid(target_datasets[member], new_lats, new_lons)
