def colReadCSV(configFile, dataFile, source):
"""
Loads a csv file containing 'column' data into a record (numpy)
array with columns labelled by 'fields'. There must be a section in
the ``configFile`` named ``source`` that sets out the format of the
data file.
:param str configFile: Path to a configuration file that holds details
of the input data.
:param str dataFile: Path to the input file to load.
:param str source: Name of the source format. There must be a
corresponding section in the ``configFile``.
:returns: A :class:`numpy.ndarray` that contains the input data.
"""
config = ConfigParser()
config.read(configFile)
delimiter = config.get(source, 'FieldDelimiter')
numHeadingLines = config.getint(source, 'NumberOfHeadingLines')
cols = config.get(source, 'Columns').split(delimiter)
usecols = [i for i,c in enumerate(cols) if c != 'skip']
data = np.genfromtxt(dataFile, dtype=None, delimiter=delimiter,
usecols=usecols, comments=None, skip_header=numHeadingLines,
autostrip=True)
data.dtype.names = [c for c in cols if c != 'skip']
return data
def historic(self):
"""Calculate historical rates of longitude crossing"""
log.info("Processing historical tracks for longitude crossings")
config = ConfigParser()
config.read(self.configFile)
inputFile = config.get("DataProcess", "InputFile")
source = config.get("DataProcess", "Source")
timestep = config.getfloat("TrackGenerator", "Timestep")
if len(os.path.dirname(inputFile)) == 0:
inputFile = pjoin(self.inputPath, inputFile)
try:
tracks = interpolateTracks.parseTracks(
self.configFile, inputFile, source, timestep, interpolation_type="linear"
)
except (TypeError, IOError, ValueError):
log.critical("Cannot load historical track file: {0}".format(inputFile))
raise
else:
self.lonCrossingHist, self.lonCrossingEWHist, self.lonCrossingWEHist = self.findCrossings(tracks)
return
def historic(self):
"""
Load historic data and calculate histogram.
Note that the input historical data is filtered by year
when it's loaded in `interpolateTracks.parseTracks()`.
The timestep to interpolate to is set to match that of the
synthetic event set (normally set to 1 hour).
"""
config = ConfigParser()
config.read(self.configFile)
inputFile = config.get('DataProcess', 'InputFile')
if len(os.path.dirname(inputFile)) == 0:
inputFile = pjoin(self.inputPath, inputFile)
source = config.get('DataProcess', 'Source')
try:
tracks = loadTrackFile(self.configFile, inputFile, source)
except (TypeError, IOError, ValueError):
log.critical("Cannot load historical track file: {0}".\
format(inputFile))
raise
else:
startYr = 9999
endYr = 0
for t in tracks:
startYr = min(startYr, min(t.Year))
endYr = max(endYr, max(t.Year))
numYears = endYr - startYr
self.hist = self.calculate(tracks) / numYears
def historic(self):
"""Load historic data and calculate histogram"""
log.info("Processing historic track records")
config = ConfigParser()
config.read(self.configFile)
inputFile = config.get('DataProcess', 'InputFile')
if len(os.path.dirname(inputFile)) == 0:
inputFile = pjoin(self.inputPath, inputFile)
source = config.get('DataProcess', 'Source')
try:
tracks = loadTrackFile(self.configFile, inputFile, source)
except (TypeError, IOError, ValueError):
log.critical("Cannot load historical track file: {0}".\
format(inputFile))
raise
else:
startYr = 9999
endYr = 0
for t in tracks:
startYr = min(startYr, min(t.Year))
endYr = max(endYr, max(t.Year))
numYears = endYr - startYr
log.info("Range of years: %d - %d" % (startYr, endYr))
try:
self.hist = self._calculate(tracks)
#self.hist = self._calculate(tracks) / numYears
except (ValueError):
log.critical(
"KDE error: The number of observations must be larger than the number of variables"
)
raise
def main(configFile):
config = ConfigParser()
config.read(configFile)
doOutputDirectoryCreation(configFile)
trackFile = config.get('DataProcess', 'InputFile')
source = config.get('DataProcess', 'Source')
delta = 1/12.
outputPath = pjoin(config.get('Output','Path'), 'tracks')
outputTrackFile = pjoin(outputPath, "tracks.interp.csv")
# This will save interpolated track data in TCRM format:
interpTrack = interpolateTracks.parseTracks(configFile, trackFile, source, delta,
outputTrackFile)
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Calculating wind fields: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import wind
wind.run(configFile, status)
# FIXME: Add wind field and timeseries plotting
doTimeseriesPlotting(configFile)
def historic(self):
"""Load historic data and calculate histogram"""
config = ConfigParser()
config.read(self.configFile)
inputFile = config.get('DataProcess', 'InputFile')
if len(os.path.dirname(inputFile)) == 0:
inputFile = pjoin(self.inputPath, inputFile)
source = config.get('DataProcess', 'Source')
try:
tracks = loadTrackFile(self.configFile, inputFile,source)
except (TypeError, IOError, ValueError):
log.critical("Cannot load historical track file: {0}".format(inputFile))
raise
else:
startYr = 9999
endYr = 0
for t in tracks:
startYr = min(startYr, min(t.Year))
endYr = max(endYr, max(t.Year))
numYears = endYr - startYr
log.info("Range of years: %d - %d" % (startYr, endYr))
self.hist = self._calculate(tracks) / numYears
def historic(self):
"""Calculate historical rates of longitude crossing"""
LOG.info("Processing historical tracks for longitude crossings")
config = ConfigParser()
config.read(self.configFile)
inputFile = config.get('DataProcess', 'InputFile')
source = config.get('DataProcess', 'Source')
timestep = config.getfloat('TrackGenerator', 'Timestep')
if len(os.path.dirname(inputFile)) == 0:
inputFile = pjoin(self.inputPath, inputFile)
try:
tracks = interpolateTracks.parseTracks(self.configFile,
inputFile,
source,
timestep,
interpolation_type='linear')
except (TypeError, IOError, ValueError):
LOG.critical("Cannot load historical track file: {0}".\
format(inputFile))
raise
else:
self.lonCrossingHist, self.lonCrossingEWHist, \
self.lonCrossingWEHist = self.findCrossings(tracks)
return
def __init__(self, configFile):
config = ConfigParser()
config.read(configFile)
self.configFile = configFile
outputPath = config.get('Output', 'Path')
self.trackPath = pjoin(outputPath, 'tracks')
self.plotPath = pjoin(outputPath, 'plots', 'stats')
self.dataPath = pjoin(outputPath, 'process')
# Determine TCRM input directory
tcrm_dir = pathLocator.getRootDirectory()
self.inputPath = pjoin(tcrm_dir, 'input')
self.synNumYears = config.getint('TrackGenerator',
'yearspersimulation')
try:
gateFile = config.get('Input', 'CoastlineGates')
except NoOptionError:
log.exception(("No coastline gate file specified "
"in configuration file"))
raise
gateData = np.genfromtxt(gateFile, delimiter=',')
nGates = len(gateData)
self.gates = Int.convert2vertex(gateData[:, 1], gateData[:, 2])
self.coast = list(self.gates)
self.coast.append(self.gates[0])
def historic(self):
"""Load historic data and calculate histogram"""
config = ConfigParser()
config.read(self.configFile)
inputFile = config.get('DataProcess', 'InputFile')
if len(os.path.dirname(inputFile)) == 0:
inputFile = pjoin(self.inputPath, inputFile)
source = config.get('DataProcess', 'Source')
timestep = config.getfloat('TrackGenerator', 'Timestep')
interpHistFile = pjoin(self.inputPath, "interp_tracks.csv")
try:
tracks = interpolateTracks.parseTracks(self.configFile,
inputFile,
source,
timestep,
interpHistFile, 'linear')
except (TypeError, IOError, ValueError):
log.critical("Cannot load historical track file: {0}".format(inputFile))
raise
else:
startYr = 9999
endYr = 0
for t in tracks:
startYr = min(startYr, min(t.Year))
endYr = max(endYr, max(t.Year))
numYears = endYr - startYr
self.hist = self.calculate(tracks) / numYears
def __init__(self, configFile):
config = ConfigParser()
config.read(configFile)
self.configFile = configFile
outputPath = config.get('Output', 'Path')
self.trackPath = pjoin(outputPath, 'tracks')
self.plotPath = pjoin(outputPath, 'plots', 'stats')
self.dataPath = pjoin(outputPath, 'process')
# Determine TCRM input directory
tcrm_dir = pathLocator.getRootDirectory()
self.inputPath = pjoin(tcrm_dir, 'input')
self.synNumYears = config.getint('TrackGenerator',
'yearspersimulation')
try:
gateFile = config.get('Input', 'CoastlineGates')
except NoOptionError:
LOG.exception(("No coastline gate file specified "
"in configuration file"))
raise
gateData = np.genfromtxt(gateFile, delimiter=',')
self.gates = Int.convert2vertex(gateData[:, 1], gateData[:, 2])
self.coast = list(self.gates)
self.coast.append(self.gates[0])
def main(configFile):
"""
Main function to execute the :mod:`wind`.
:param str configFile: Path to configuration file.
"""
config = ConfigParser()
config.read(configFile)
doOutputDirectoryCreation(configFile)
trackFile = config.get('DataProcess', 'InputFile')
source = config.get('DataProcess', 'Source')
delta = 1/12.
outputPath = pjoin(config.get('Output','Path'), 'tracks')
outputTrackFile = pjoin(outputPath, "tracks.interp.csv")
# This will save interpolated track data in TCRM format:
interpTrack = interpolateTracks.parseTracks(configFile, trackFile,
source, delta,
outputTrackFile,
interpolation_type='akima')
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Calculating wind fields: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import wind
wind.run(configFile, status)
doWindfieldPlotting(configFile)
doTimeseriesPlotting(configFile)
def __init__(self, configFile):
config = ConfigParser()
config.read(configFile)
self.outputPath = config.get('Output', 'Path')
self.windfieldPath = pjoin(self.outputPath, 'windfield')
self.trackPath = pjoin(self.outputPath, 'tracks')
self.hazardPath = pjoin(self.outputPath, 'hazard')
self.domain = config.geteval('Region', 'gridLimit')
self.hazardDB = pjoin(self.outputPath, 'hazard.db')
self.locationDB = pjoin(self.outputPath, 'locations.db')
self.datfile = config.get('Process', 'DatFile')
self.excludePastProcessed = config.getboolean('Process',
'ExcludePastProcessed')
pGetProcessedFiles(self.datfile)
sqlite3.Connection.__init__(self,
self.hazardDB,
detect_types=PARSE_DECLTYPES
| PARSE_COLNAMES)
self.exists = True
import atexit
atexit.register(self.close)
def startup():
"""
Parse command line arguments and call the :func:`main` function.
"""
parser = argparse.ArgumentParser()
parser.add_argument('-c',
'--config_file',
help='Path to configuration file')
parser.add_argument('-v',
'--verbose',
help='Verbose output',
action='store_true')
parser.add_argument('-d',
'--debug',
help='Allow pdb traces',
action='store_true')
args = parser.parse_args()
configFile = args.config_file
config = ConfigParser()
config.read(configFile)
rootdir = pathLocator.getRootDirectory()
os.chdir(rootdir)
logfile = config.get('Logging', 'LogFile')
logdir = dirname(realpath(logfile))
# If log file directory does not exist, create it
if not isdir(logdir):
try:
os.makedirs(logdir)
except OSError:
logfile = pjoin(os.getcwd(), 'processMultipliers.log')
logLevel = config.get('Logging', 'LogLevel')
verbose = config.getboolean('Logging', 'Verbose')
datestamp = config.getboolean('Logging', 'Datestamp')
debug = False
if args.verbose:
verbose = True
if args.debug:
debug = True
flStartLog(logfile, logLevel, verbose, datestamp)
if debug:
main(configFile)
else:
try:
modified_main(configFile)
except Exception: # pylint: disable=W0703
# Catch any exceptions that occur and log them (nicely):
tblines = traceback.format_exc().splitlines()
for line in tblines:
log.critical(line.lstrip())
def doWindfieldPlotting(configFile):
"""
Plot the wind field on a map.
:param str configFile: Path to the configuration file.
:Note: the file name is assumed to be 'gust.interp.nc'
"""
from netCDF4 import Dataset
import numpy as np
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output', 'Path')
windfieldPath = pjoin(outputPath, 'windfield')
inputFile = config.get('DataProcess', 'InputFile')
if inputFile.endswith(".nc"):
# We have a netcdf track file. Work under the assumption it is
# drawn directly from TCRM.
trackFile = os.path.basename(inputFile)
trackId = trackFile.split('.')[1]
gustFile = 'gust.{0}.nc'.format(trackId)
outputWindFile = pjoin(windfieldPath, gustFile)
else:
# Note the assumption about the file name!
outputWindFile = pjoin(windfieldPath, 'gust.001-00001.nc')
plotPath = pjoin(outputPath, 'plots', 'maxwind.png')
f = Dataset(outputWindFile, 'r')
xdata = f.variables['lon'][:]
ydata = f.variables['lat'][:]
vdata = f.variables['vmax'][:]
gridLimit = None
if config.has_option('Region', 'gridLimit'):
gridLimit = config.geteval('Region', 'gridLimit')
ii = np.where((xdata >= gridLimit['xMin'])
& (xdata <= gridLimit['xMax']))
jj = np.where((ydata >= gridLimit['yMin'])
& (ydata <= gridLimit['yMax']))
[xgrid, ygrid] = np.meshgrid(xdata[ii], ydata[jj])
ig, jg = np.meshgrid(ii, jj)
vdata = vdata[jg, ig]
else:
[xgrid, ygrid] = np.meshgrid(xdata, ydata)
map_kwargs = dict(llcrnrlon=xgrid.min(),
llcrnrlat=ygrid.min(),
urcrnrlon=xgrid.max(),
urcrnrlat=ygrid.max(),
projection='merc',
resolution='i')
title = "Maximum wind speed"
cbarlabel = "Wind speed ({0})".format(f.variables['vmax'].units)
levels = np.arange(30, 101., 5.)
saveWindfieldMap(vdata, xgrid, ygrid, title, levels, cbarlabel, map_kwargs,
plotPath)
def __init__(self, configFile, tilegrid, numSim, minRecords, yrsPerSim,
calcCI=False, evd='GEV'):
"""
Initialise HazardCalculator object.
:param str configFile: path to TCRM configuration file.
:param tilegrid: :class:`TileGrid` instance
:param int numSim: number of simulations created.
:param int minRecords: minimum number of valid wind speed values required
to do fitting.
:param int yrsPerSim:
:param boolean calcCI:
:param str extreme_value_distribution: evd to use. Options so far are GEV
and GPD.
"""
config = ConfigParser()
config.read(configFile)
self.nodata = -9999.
self.years = np.array(config.get('Hazard',
'Years').split(',')).astype('f')
self.outputPath = pjoin(config.get('Output', 'Path'), 'hazard')
self.inputPath = pjoin(config.get('Output', 'Path'), 'windfield')
gridLimit = config.geteval('Region', 'gridLimit')
self.numSim = numSim
self.minRecords = minRecords
self.yrsPerSim = yrsPerSim
self.calcCI = calcCI
if self.calcCI:
log.debug("Bootstrap confidence intervals will be calculated")
self.sample_size = config.getint('Hazard', 'SampleSize')
self.prange = config.getint('Hazard', 'PercentileRange')
self.evd = evd
self.tilegrid = tilegrid
lon, lat = self.tilegrid.getDomainExtent()
# Create arrays for storing output data:
self.loc = np.zeros((len(lat), len(lon)), dtype='f')
self.shp = np.zeros((len(lat), len(lon)), dtype='f')
self.scale = np.zeros((len(lat), len(lon)), dtype='f')
self.Rp = np.zeros((len(self.years), len(lat), len(lon)), dtype='f')
self.RPupper = np.zeros((len(self.years), len(lat), len(lon)), dtype='f')
self.RPlower = np.zeros((len(self.years), len(lat), len(lon)), dtype='f')
self.global_atts = {'title': ('TCRM hazard simulation - '
'return period wind speeds'),
'tcrm_version': flProgramVersion(),
'python_version': sys.version}
# Add configuration settings to global attributes:
for section in config.sections():
for option in config.options(section):
key = "{0}_{1}".format(section, option)
value = config.get(section, option)
self.global_atts[key] = value
def startup():
parser = argparse.ArgumentParser()
parser.add_argument('-c', '--config_file',
help='Path to configuration file')
parser.add_argument('-v', '--verbose', help='Verbose output',
action='store_true')
parser.add_argument('-d', '--debug', help='Allow pdb traces',
action='store_true')
args = parser.parse_args()
configFile = args.config_file
config = ConfigParser()
config.read(configFile)
rootdir = pathLocator.getRootDirectory()
os.chdir(rootdir)
logfile = config.get('Logging','LogFile')
logdir = dirname(realpath(logfile))
# If log file directory does not exist, create it
if not isdir(logdir):
try:
os.makedirs(logdir)
except OSError:
logfile = pjoin(os.getcwd(), 'tcrm.log')
logLevel = config.get('Logging', 'LogLevel')
verbose = config.getboolean('Logging', 'Verbose')
debug = False
if args.verbose:
verbose = True
if args.debug:
debug = True
flStartLog(logfile, logLevel, verbose)
# Switch off minor warning messages
import warnings
warnings.filterwarnings("ignore", category=DeprecationWarning)
warnings.filterwarnings("ignore", category=UserWarning, module="pytz")
warnings.filterwarnings("ignore", category=UserWarning, module="numpy")
warnings.filterwarnings("ignore", category=UserWarning,
module="matplotlib")
warnings.filterwarnings("ignore", category=RuntimeWarning)
if debug:
main(configFile)
else:
try:
main(configFile)
except Exception: # pylint: disable=W0703
# Catch any exceptions that occur and log them (nicely):
tblines = traceback.format_exc().splitlines()
for line in tblines:
log.critical(line.lstrip())
def run(configFile, callback=None):
"""
Run the hazard calculations.
This will attempt to run the calculation in parallel by tiling the
domain, but also provides a sane fallback mechanism to execute
in serial.
:param str configFile: path to configuration file
"""
log.info("Loading hazard calculation settings")
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output', 'Path')
inputPath = pjoin(outputPath, 'windfield')
gridLimit = config.geteval('Region', 'gridLimit')
numsimulations = config.getint('TrackGenerator', 'NumSimulations')
yrsPerSim = config.getint('TrackGenerator', 'YearsPerSimulation')
minRecords = config.getint('Hazard', 'MinimumRecords')
calculate_confidence = config.getboolean('Hazard', 'CalculateCI')
extreme_value_distribution = config.get('Hazard', 'ExtremeValueDistribution')
wf_lon, wf_lat = setDomain(inputPath)
global MPI, comm
MPI = attemptParallel()
comm = MPI.COMM_WORLD
log.info("Running hazard calculations")
TG = TileGrid(gridLimit, wf_lon, wf_lat)
tiles = getTiles(TG)
#def progress(i):
# callback(i, len(tiles))
comm.barrier()
hc = HazardCalculator(configFile, TG,
numsimulations,
minRecords,
yrsPerSim,
calculate_confidence,
extreme_value_distribution
)
hc.dumpHazardFromTiles(tiles)
log.debug("Finished hazard calculations")
comm.barrier()
hc.saveHazard()
log.info("Completed hazard calculation")
def modified_main(config_file):
"""
Main function to combine the multipliers with the regional wind
speed data.
:param str configFile: Path to configuration file.
"""
config = ConfigParser()
config.read(config_file)
input_path = config.get('Input', 'Path')
try:
gust_file = config.get('Input', 'Gust_file')
except:
gust_file = 'gust.001-00001.nc'
windfield_path = pjoin(input_path, 'windfield')
ncfile = pjoin(windfield_path, gust_file)
multiplier_path = config.get('Input', 'Multipliers')
# Load the wind data:
log.info("Loading regional wind data from {0}".format(ncfile))
ncobj = Dataset(ncfile, 'r')
lat = ncobj.variables['lat'][:]
lon = ncobj.variables['lon'][:]
delta = lon[1] - lon[0]
lon = lon - delta / 2.
lat = lat - delta / 2.
# Wind speed:
wspd = ncobj.variables['vmax'][:]
# Components:
uu = ncobj.variables['ua'][:]
vv = ncobj.variables['va'][:]
bearing = calculateBearing(uu, vv)
gust = wspd
Vx = uu
Vy = vv
P = None
# WARNING, THESE COULD BE WRONG!!!
# Plus it doesn't do anything,
# except hightlight these var's are going in..
lon = lon
lat = lat
# Need to be checked !!!
# Load a multiplier file to determine the projection:
log.info("Using M4 data from {0}".format(multiplier_path))
processMult(gust, Vx, Vy, lon, lat, windfield_path, multiplier_path)
def __init__(self, configFile, tilegrid, numSim, minRecords, yrsPerSim,
calcCI=False):
"""
Initialise HazardCalculator object.
:param str configFile: path to TCRM configuration file.
:param tilegrid: :class:`TileGrid` instance
:param int numSim: number of simulations created.
:param int minRecords: minimum number of valid wind speed values required
to do fitting.
:param int yrsPerSim:
"""
config = ConfigParser()
config.read(configFile)
self.nodata = -9999.
self.years = np.array(config.get('Hazard',
'Years').split(',')).astype('f')
self.outputPath = pjoin(config.get('Output', 'Path'), 'hazard')
self.inputPath = pjoin(config.get('Output', 'Path'), 'windfield')
gridLimit = config.geteval('Region', 'gridLimit')
self.numSim = numSim
self.minRecords = minRecords
self.yrsPerSim = yrsPerSim
self.calcCI = calcCI
if self.calcCI:
log.debug("Bootstrap confidence intervals will be calculated")
self.sample_size = config.getint('Hazard', 'SampleSize')
self.prange = config.getint('Hazard', 'PercentileRange')
self.tilegrid = tilegrid
lon, lat = self.tilegrid.getDomainExtent()
# Create arrays for storing output data:
self.loc = np.zeros((len(lat), len(lon)), dtype='f')
self.shp = np.zeros((len(lat), len(lon)), dtype='f')
self.scale = np.zeros((len(lat), len(lon)), dtype='f')
self.Rp = np.zeros((len(self.years), len(lat), len(lon)), dtype='f')
self.RPupper = np.zeros((len(self.years), len(lat), len(lon)), dtype='f')
self.RPlower = np.zeros((len(self.years), len(lat), len(lon)), dtype='f')
self.global_atts = {'history': ('TCRM hazard simulation - '
'return period wind speeds'),
'version': flProgramVersion(),
'Python_ver': sys.version}
# Add configuration settings to global attributes:
for section in config.sections():
for option in config.options(section):
key = "{0}_{1}".format(section, option)
value = config.get(section, option)
self.global_atts[key] = value
def startup():
"""
Parse command line arguments and call the :func:`main` function.
"""
parser = argparse.ArgumentParser()
parser.add_argument('-c', '--config_file',
help='Path to configuration file')
parser.add_argument('-v', '--verbose', help='Verbose output',
action='store_true')
parser.add_argument('-d', '--debug', help='Allow pdb traces',
action='store_true')
args = parser.parse_args()
config_file = args.config_file
config = ConfigParser()
config.read(config_file)
rootdir = pathLocator.getRootDirectory()
os.chdir(rootdir)
logfile = config.get('Logging', 'LogFile')
logdir = dirname(realpath(logfile))
# If log file directory does not exist, create it
if not isdir(logdir):
try:
os.makedirs(logdir)
except OSError:
logfile = pjoin(os.getcwd(), 'tsmultipliers.log')
logLevel = config.get('Logging', 'LogLevel')
verbose = config.getboolean('Logging', 'Verbose')
datestamp = config.getboolean('Logging', 'Datestamp')
debug = False
if args.verbose:
verbose = True
if args.debug:
debug = True
flStartLog(logfile, logLevel, verbose, datestamp)
if debug:
process_timeseries(config_file)
else:
try:
process_timeseries(config_file)
except Exception: # pylint: disable=W0703
# Catch any exceptions that occur and log them (nicely):
tblines = traceback.format_exc().splitlines()
for line in tblines:
log.critical(line.lstrip())
log.info("Completed {0}".format(sys.argv[0]))
def __init__(self, configFile):
"""
Read configuration settings, load station data and set up
output recarrays.
:param str configFile: path to a configuration file.
"""
config = ConfigParser()
config.read(configFile)
self.meta = False
stnFile = config.get('Timeseries', 'LocationFile')
self.outputPath = pjoin(config.get('Output', 'Path'),
'process', 'timeseries')
self.maxfile = pjoin(config.get('Output', 'Path'),
'process', 'maxima.csv')
self.minfile = pjoin(config.get('Output', 'Path'),
'process', 'minima.csv')
log.info(f"Loading timeseries stations from {stnFile}")
log.debug(f"Timeseries data will be written into {self.outputPath}")
self.stations = []
if stnFile.endswith("shp"):
try:
key_name = config.get('Timeseries', 'StationID')
except NoOptionError:
key_name = None
vertices = shpGetVertices(stnFile, key_name=key_name)
for stn in list(vertices.keys()):
lat = vertices[stn][0][1]
lon = vertices[stn][0][0]
lon = np.where(lon < 0., lon + 360., lon)
self.stations.append(Station(stn, lon, lat))
else:
stndata = flLoadFile(stnFile, delimiter=',')
# If there are more than 3 columns, save the additional
# columns as 'metadata'
if stndata.shape[1] > 3:
self.metadata = stndata[:, 3:]
self.meta = True
stnid = stndata[:, 0]
stnlon = stndata[:, 1].astype(float)
stnlat = stndata[:, 2].astype(float)
for sid, lon, lat in zip(stnid, stnlon, stnlat):
self.stations.append(Station(sid, lon, lat))
log.info(f"There are {len(self.stations)} stations that will collect timeseries data")
def __init__(self, configFile, progressbar=None):
"""
Initialize the data include tool instance, Nan value and all
full path names of the files in which data will be stored.
"""
#CalcTD = CalcTrackDomain(configFile)
#self.domain = CalcTD.calc()
self.configFile = configFile
self.progressbar = progressbar
self.logger = logging.getLogger(__name__)
self.logger.info("Initialising DataProcess")
config = ConfigParser()
config.read(configFile)
self.outputPath = config.get('Output', 'Path')
self.processPath = pjoin(self.outputPath, 'process')
# Determine TCRM input directory
tcrm_dir = pathLocator.getRootDirectory()
self.tcrm_input_dir = pjoin(tcrm_dir, 'input')
landmask = config.get('Input', 'LandMask')
self.landmask = SampleGrid(landmask)
fmt = config.get('Output', 'Format')
self.ncflag = False
if fmt.startswith("nc"):
self.logger.debug("Output format is netcdf")
self.ncflag = True
self.data = {}
#dimensions = {records}
# variables = {init_index(records),
# genesis_index(records),
# non_init_index(records),
# lon(records), lat(records),
# year(records), month(records),
# day(records), hour(records),
# minute(records), julianday(records),
# bearing(records), speed(records),
# pressure(records), lsflag(records), }
# global_attributes = dict(description=
# source_file=,
# source_file_moddate,
# landmask_file=,
# version=,)
elif fmt.startswith("txt"):
self.logger.debug("Output format is text")
self.origin_year = pjoin(self.processPath, 'origin_year')
def __init__(self, configFile):
"""
Read configuration settings, load station data and set up
output recarrays.
:param str configFile: path to a configuration file.
"""
config = ConfigParser()
config.read(configFile)
self.meta = False
stnFile = config.get('Timeseries', 'StationFile')
self.outputPath = pjoin(config.get('Output', 'Path'),
'process', 'timeseries')
self.maxfile = pjoin(config.get('Output', 'Path'),
'process', 'maxima.csv')
self.minfile = pjoin(config.get('Output', 'Path'),
'process', 'minima.csv')
log.debug("Loading stations from %s"%stnFile)
log.debug("Timeseries data will be written into %s"%self.outputPath)
self.stations = []
if stnFile.endswith("shp"):
try:
key_name = config.get('Timeseries', 'StationID')
except NoOptionError:
key_name = None
vertices = shpGetVertices(stnFile, key_name=key_name)
for stn in vertices.keys():
self.stations.append(Station(stn, vertices[stn][0][0],
vertices[stn][0][1]))
else:
stndata = flLoadFile(stnFile, delimiter=',')
# If there are more than 3 columns, save the additional
# columns as 'metadata'
if stndata.shape[1] > 3:
self.metadata = stndata[:, 3:]
self.meta = True
stnid = stndata[:, 0]
stnlon = stndata[:, 1].astype(float)
stnlat = stndata[:, 2].astype(float)
for sid, lon, lat in zip(stnid, stnlon, stnlat):
self.stations.append(Station(sid, lon, lat))
def loadDatasets():
config = ConfigParser()
datasets = config.get('Input', 'Datasets').split(',')
global DATASETS
for dataset in datasets:
url = config.get(dataset, 'URL')
path = config.get(dataset, 'path')
if config.has_option(dataset, 'filename'):
filename = config.get(dataset, 'filename')
else:
filename = None
data = DataSet(dataset, url, path, filename)
DATASETS.append(data)
def __init__(self, configFile, autoCalc_gridLimit=None,
progressbar=None):
"""
Initialize the data and variables required for the interface
"""
self.configFile = configFile
config = ConfigParser()
config.read(configFile)
self.progressbar = progressbar
log.info("Initialising StatInterface")
self.kdeType = config.get('StatInterface', 'kdeType')
self.kde2DType = config.get('StatInterface','kde2DType')
minSamplesCell = config.getint('StatInterface', 'minSamplesCell')
self.kdeStep = config.getfloat('StatInterface', 'kdeStep')
self.outputPath = config.get('Output', 'Path')
self.processPath = pjoin(self.outputPath, 'process')
missingValue = cnfGetIniValue(self.configFile, 'StatInterface',
'MissingValue', sys.maxint)
gridLimitStr = cnfGetIniValue(self.configFile, 'StatInterface',
'gridLimit', '')
if gridLimitStr is not '':
try:
self.gridLimit = eval(gridLimitStr)
except SyntaxError:
log.exception('Error! gridLimit is not a dictionary')
else:
self.gridLimit = autoCalc_gridLimit
log.info('No gridLimit specified - using automatic' +
' selection: ' + str(self.gridLimit))
try:
gridSpace = config.geteval('Region', 'gridSpace')
gridInc = config.geteval('Region', 'gridInc')
except SyntaxError:
log.exception('Error! gridSpace or gridInc not dictionaries')
raise
self.generateDist = GenerateDistributions(self.configFile,
self.gridLimit,
gridSpace, gridInc,
self.kdeType,
minSamplesCell,
missingValue)
self.gridSpace = gridSpace
self.gridInc = gridInc
def __init__(self, configFile, progressbar=None):
#CalcTD = CalcTrackDomain(configFile)
#self.domain = CalcTD.calc()
self.configFile = configFile
self.progressbar = progressbar
self.logger = logging.getLogger(__name__)
self.logger.info("Initialising DataProcess")
config = ConfigParser()
config.read(configFile)
self.outputPath = config.get('Output', 'Path')
self.processPath = pjoin(self.outputPath, 'process')
# Determine TCRM input directory
tcrm_dir = pathLocator.getRootDirectory()
self.tcrm_input_dir = pjoin(tcrm_dir, 'input')
landmask = config.get('Input', 'LandMask')
self.landmask = SampleGrid(landmask)
fmt = config.get('Output', 'Format')
self.ncflag = False
if fmt.startswith("nc"):
self.logger.debug("Output format is netcdf")
self.ncflag = True
self.data = {}
#dimensions = {records}
# variables = {init_index(records),
# genesis_index(records),
# non_init_index(records),
# lon(records), lat(records),
# year(records), month(records),
# day(records), hour(records),
# minute(records), julianday(records),
# bearing(records), speed(records),
# pressure(records), lsflag(records), }
# global_attributes = dict(description=
# source_file=,
# source_file_moddate,
# landmask_file=,
# version=,)
elif fmt.startswith("txt"):
self.logger.debug("Output format is text")
self.origin_year = pjoin(self.processPath, 'origin_year')
def doHazard(configFile):
"""
Do the hazard calculations (extreme value distribution fitting)
using the :mod:`hazard` module.
:param str configFile: Name of configuration file.
"""
log.info('Running HazardInterface')
config = ConfigParser()
config.read(configFile)
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Performing hazard calculations: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import hazard
hazard.run(configFile)
log.info('Completed HazardInterface')
pbar.update(1.0)
def doHazardPlotting(configFile):
"""
Do the hazard plots (hazard maps and curves for all locations within
the model domain). Plotting is performed by the
:mod:`PlotInterface.AutoPlotHazard` module.
:param str configFile: Name of configuration file.
"""
import matplotlib
matplotlib.use('Agg') # Use matplotlib backend
config = ConfigParser()
config.read(configFile)
log.info('Plotting Hazard Maps')
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Plotting hazard maps: ', showProgressBar)
pbar.update(0.0)
from PlotInterface.AutoPlotHazard import AutoPlotHazard
plotter = AutoPlotHazard(configFile, progressbar=pbar)
plotter.plotMap()
plotter.plotCurves()
pbar.update(1.0)
def doDataDownload(configFile):
"""
Check and download the data files.
:param str configFile: Name of configuration file.
"""
log.info('Checking availability of input data sets')
config = ConfigParser()
config.read(configFile)
showProgressBar = config.get('Logging', 'ProgressBar')
for dataset in datasets.DATASETS:
if not dataset.isDownloaded():
log.info('Input file %s is not available', dataset.filename)
try:
log.info('Attempting to download %s', dataset.filename)
pbar = ProgressBar('Downloading file %s: ' % dataset.filename,
showProgressBar)
def status(fn, done, size):
pbar.update(float(done)/size)
dataset.download(status)
log.info('Download successful')
except IOError:
log.error('Unable to download %s. Maybe a proxy problem?',
dataset.filename)
sys.exit(1)
def doTrackGeneration(configFile):
"""
Do the tropical cyclone track generation.
The track generation settings are read from *configFile*.
:param str configFile: Name of configuration file.
"""
log.info('Starting track generation')
config = ConfigParser()
config.read(configFile)
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Simulating cyclone tracks: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import TrackGenerator
TrackGenerator.run(configFile, status)
pbar.update(1.0)
log.info('Completed track generation')
def doWindfieldCalculations(configFile):
"""
Do the wind field calculations. The wind field settings are read
from *configFile*.
:param str configFile: Name of configuration file.
"""
log.info('Starting wind field calculations')
config = ConfigParser()
config.read(configFile)
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Calculating wind fields: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import wind
wind.run(configFile, status)
pbar.update(1.0)
log.info('Completed wind field calculations')
def doHazardPlotting(configFile):
"""
Do the hazard plots.
:param str configFile: Name of configuration file.
"""
import matplotlib
matplotlib.use('Agg') # Use matplotlib backend
config = ConfigParser()
config.read(configFile)
log.info('Plotting Hazard Maps')
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Plotting hazard maps: ', showProgressBar)
pbar.update(0.0)
from PlotInterface.AutoPlotHazard import AutoPlotHazard
plotter = AutoPlotHazard(configFile, progressbar=pbar)
plotter.plotMap()
plotter.plotCurves()
pbar.update(1.0)
def __init__(self, configFile, kdeType, gridLimit, kdeStep, lonLat=None, progressbar=None):
"""
"""
self.logger = logging.getLogger()
self.progressbar = progressbar
if self.progressbar:
KPDF.set_callback(self.updateProgressBar)
self.logger.info("Initialising KDEOrigins")
self.configFile = configFile
self.x = numpy.arange(gridLimit['xMin'], gridLimit['xMax'], kdeStep)
self.y = numpy.arange(gridLimit['yMax'], gridLimit['yMin'], -kdeStep)
self.kdeType = kdeType
self.kdeStep = kdeStep
config = ConfigParser()
config.read(configFile)
if lonLat is None:
self.outputPath = config.get('Output', 'Path')
self.processPath = os.path.join(self.outputPath, 'process')
self.logger.debug("Loading "+os.path.join(self.processPath,
'init_lon_lat'))
ll = flLoadFile(os.path.join(self.processPath, 'init_lon_lat'),
'%', ',')
self.lonLat = ll[:,0:2]
else:
self.lonLat = lonLat[:,0:2]
self.bw = KPDF.MPDFOptimumBandwidth(self.lonLat)
self.logger.debug("Optimal bandwidth: %f"%self.bw)
def doHazardPlotting(configFile):
"""
Do the hazard plots (hazard maps and curves for all locations within
the model domain). Plotting is performed by the
:mod:`PlotInterface.AutoPlotHazard` module.
:param str configFile: Name of configuration file.
"""
config = ConfigParser()
config.read(configFile)
log.info('Plotting Hazard Maps')
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Plotting hazard maps: ', showProgressBar)
pbar.update(0.0)
from PlotInterface.AutoPlotHazard import AutoPlotHazard
plotter = AutoPlotHazard(configFile, progressbar=pbar)
plotter.plotMap()
plotter.plotCurves()
pbar.update(1.0)
def doHazard(configFile):
"""
Do the hazard calculations (extreme value distribution fitting)
using the :mod:`hazard` module.
:param str configFile: Name of configuration file.
"""
log.info('Running HazardInterface')
config = ConfigParser()
config.read(configFile)
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Performing hazard calculations: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import hazard
hazard.run(configFile)
log.info('Completed HazardInterface')
pbar.update(1.0)
def __init__(self, configFile):
"""
Calculate density of TC positions on a grid
:param str configFile: path to a TCRM configuration file.
"""
config = ConfigParser()
config.read(configFile)
self.configFile = configFile
# Define the grid:
gridLimit = config.geteval('Region', 'gridLimit')
gridSpace = config.geteval('Region', 'GridSpace')
self.lon_range = np.arange(gridLimit['xMin'],
gridLimit['xMax'] + 0.1,
gridSpace['x'])
self.lat_range = np.arange(gridLimit['yMin'],
gridLimit['yMax'] + 0.1,
gridSpace['y'])
outputPath = config.get('Output', 'Path')
self.trackPath = pjoin(outputPath, 'tracks')
self.plotPath = pjoin(outputPath, 'plots', 'stats')
self.dataPath = pjoin(outputPath, 'process')
# Determine TCRM input directory
tcrm_dir = pathLocator.getRootDirectory()
self.inputPath = pjoin(tcrm_dir, 'input')
self.synNumYears = config.getint('TrackGenerator',
'yearspersimulation')
def __init__(self, configFile):
"""
Calculate density of TC genesis positions on a grid
:param str configFile: path to a TCRM configuration file.
"""
config = ConfigParser()
config.read(configFile)
self.configFile = configFile
# Define the grid:
self.gridLimit = config.geteval('Region', 'gridLimit')
self.gridSpace = config.geteval('Region', 'GridSpace')
self.lon_range = np.arange(self.gridLimit['xMin'],
self.gridLimit['xMax'] + 0.1, 1.)
self.lat_range = np.arange(self.gridLimit['yMin'],
self.gridLimit['yMax'] + 0.1, 1.)
self.X, self.Y = np.meshgrid(self.lon_range, self.lat_range)
outputPath = config.get('Output', 'Path')
self.trackPath = pjoin(outputPath, 'tracks')
self.plotPath = pjoin(outputPath, 'plots', 'stats')
self.dataPath = pjoin(outputPath, 'process')
# Determine TCRM input directory
tcrm_dir = pathLocator.getRootDirectory()
self.inputPath = pjoin(tcrm_dir, 'input')
self.synNumYears = config.getint('TrackGenerator',
'yearspersimulation')
def doOutputDirectoryCreation(configFile):
"""
Create all the necessary output folders.
:param str configFile: Name of configuration file.
:raises OSError: If the directory tree cannot be created.
"""
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output', 'Path')
log.info('Output will be stored under %s', outputPath)
subdirs = ['tracks', 'windfield', 'plots', 'plots/timeseries',
'log', 'process', 'process/timeseries']
if not isdir(outputPath):
try:
os.makedirs(outputPath)
except OSError:
raise
for subdir in subdirs:
if not isdir(realpath(pjoin(outputPath, subdir))):
try:
os.makedirs(realpath(pjoin(outputPath, subdir)))
except OSError:
raise
def doTrackGeneration(configFile):
"""
Do the tropical cyclone track generation in :mod:`TrackGenerator`.
The track generation settings are read from *configFile*.
:param str configFile: Name of configuration file.
"""
log.info('Starting track generation')
config = ConfigParser()
config.read(configFile)
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Simulating cyclone tracks: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import TrackGenerator
TrackGenerator.run(configFile, status)
pbar.update(1.0)
log.info('Completed track generation')
def doOutputDirectoryCreation(configFile):
"""
Create all the necessary output folders.
:param str configFile: Name of configuration file.
:raises OSError: If the directory tree cannot be created.
"""
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output', 'Path')
log.info('Output will be stored under %s', outputPath)
subdirs = ['tracks', 'hazard', 'windfield', 'plots', 'plots/hazard',
'plots/stats', 'log', 'process', 'process/timeseries',
'process/dat']
if not isdir(outputPath):
try:
os.makedirs(outputPath)
except (OSError, FileExistsError):
raise
for subdir in subdirs:
if not isdir(realpath(pjoin(outputPath, subdir))):
try:
os.makedirs(realpath(pjoin(outputPath, subdir)))
except (OSError, FileExistsError):
raise
def doWindfieldCalculations(configFile):
"""
Do the wind field calculations, using :mod:`wind`. The wind
field settings are read from *configFile*.
:param str configFile: Name of configuration file.
"""
log.info('Starting wind field calculations')
config = ConfigParser()
config.read(configFile)
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Calculating wind fields: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import wind
wind.run(configFile, status)
pbar.update(1.0)
log.info('Completed wind field calculations')
def __init__(self, configFile, progressbar=None):
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output', 'Path')
try:
self.localityID = config.get('Region', 'LocalityID')
except Exception:
self.localityID = -999999
self.inputFile = pjoin(outputPath, 'hazard', 'hazard.nc')
self.plotPath = pjoin(outputPath, 'plots', 'hazard')
self.plotUnits = PlotUnits(config.get('Hazard', 'PlotSpeedUnits'))
self.progressbar = progressbar
def __init__(self, configFile, autoCalc_gridLimit=None, progressbar=None):
"""
Initialize the data and variables required for the interface
"""
self.configFile = configFile
config = ConfigParser()
config.read(configFile)
self.progressbar = progressbar
log.info("Initialising StatInterface")
self.kdeType = config.get('StatInterface', 'kdeType')
minSamplesCell = config.getint('StatInterface', 'minSamplesCell')
self.kdeStep = config.getfloat('StatInterface', 'kdeStep')
self.outputPath = config.get('Output', 'Path')
self.processPath = pjoin(self.outputPath, 'process')
missingValue = cnfGetIniValue(self.configFile, 'StatInterface',
'MissingValue', sys.maxsize)
gridLimitStr = cnfGetIniValue(self.configFile, 'StatInterface',
'gridLimit', '')
if gridLimitStr is not '':
try:
self.gridLimit = eval(gridLimitStr)
except SyntaxError:
log.exception('Error! gridLimit is not a dictionary')
else:
self.gridLimit = autoCalc_gridLimit
log.info('No gridLimit specified - using automatic' +
' selection: ' + str(self.gridLimit))
try:
gridSpace = config.geteval('Region', 'gridSpace')
gridInc = config.geteval('Region', 'gridInc')
except SyntaxError:
log.exception('Error! gridSpace or gridInc not dictionaries')
raise
self.generateDist = GenerateDistributions(self.configFile,
self.gridLimit, gridSpace,
gridInc, self.kdeType,
minSamplesCell, missingValue)
self.gridSpace = gridSpace
self.gridInc = gridInc
def __init__(self, configFile, progressbar=None):
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output','Path')
try:
self.localityID = config.get('Region', 'LocalityID')
except Exception:
self.localityID = -999999
self.inputFile = pjoin(outputPath, 'hazard', 'hazard.nc')
self.plotPath = pjoin(outputPath, 'plots', 'hazard')
self.plotUnits = PlotUnits(config.get('Hazard', 'PlotSpeedUnits'))
self.progressbar = progressbar
def __init__(self, configFile):
config = ConfigParser()
config.read(configFile)
self.configFile = configFile
# Define the grid:
gridLimit = config.geteval('Region', 'gridLimit')
gridSpace = config.geteval('Region', 'GridSpace')
self.lon_range = np.arange(gridLimit['xMin'], gridLimit['xMax'] + 0.1,
gridSpace['x'])
self.lat_range = np.arange(gridLimit['yMin'], gridLimit['yMax'] + 0.1,
gridSpace['y'])
outputPath = config.get('Output', 'Path')
self.trackPath = pjoin(outputPath, 'tracks')
self.plotPath = pjoin(outputPath, 'plots', 'stats')
self.dataPath = pjoin(outputPath, 'process')
# Determine TCRM input directory
tcrm_dir = pathLocator.getRootDirectory()
self.inputPath = pjoin(tcrm_dir, 'input')
self.synNumYears = config.getint('TrackGenerator',
'yearspersimulation')
# Longitude crossing gates:
self.gateLons = np.arange(self.lon_range.min(),
self.lon_range.max() + 0.5, 10.)
self.gateLats = np.arange(self.lat_range.min(),
self.lat_range.max() + 0.5, 2.)
# Add configuration settings to global attributes:
self.gatts = {
'history': "Longitude crossing rates for TCRM simulation",
'version': flProgramVersion()
}
for section in config.sections():
for option in config.options(section):
key = "{0}_{1}".format(section, option)
value = config.get(section, option)
self.gatts[key] = value
def __init__(self, configFile):
config = ConfigParser()
config.read(configFile)
self.configFile = configFile
# Define the grid:
gridLimit = config.geteval('Region', 'gridLimit')
gridSpace = config.geteval('Region', 'GridSpace')
self.lon_range = np.arange(gridLimit['xMin'],
gridLimit['xMax'] + 0.1,
gridSpace['x'])
self.lat_range = np.arange(gridLimit['yMin'],
gridLimit['yMax'] + 0.1,
gridSpace['y'])
outputPath = config.get('Output', 'Path')
self.trackPath = pjoin(outputPath, 'tracks')
self.plotPath = pjoin(outputPath, 'plots', 'stats')
self.dataPath = pjoin(outputPath, 'process')
# Determine TCRM input directory
tcrm_dir = pathLocator.getRootDirectory()
self.inputPath = pjoin(tcrm_dir, 'input')
self.synNumYears = config.getint('TrackGenerator',
'yearspersimulation')
# Longitude crossing gates:
self.gateLons = np.arange(self.lon_range.min(),
self.lon_range.max() + 0.5, 10.)
self.gateLats = np.arange(self.lat_range.min(),
self.lat_range.max() + 0.5, 2.)
# Add configuration settings to global attributes:
self.gatts = {'history': "Longitude crossing rates for TCRM simulation",
'version': flProgramVersion() }
for section in config.sections():
for option in config.options(section):
key = "{0}_{1}".format(section, option)
value = config.get(section, option)
self.gatts[key] = value
def __init__(self, configFile):
"""
:param str configFile: Path to configuration file.
"""
config = ConfigParser()
config.read(configFile)
self.outputPath = config.get('Output', 'Path')
self.wf_domain = config.geteval('Region', 'gridLimit')
def __init__(self, configFile):
"""
:param str configFile: Path to configuration file.
"""
config = ConfigParser()
config.read(configFile)
self.outputPath = config.get('Output', 'Path')
self.wf_domain = config.geteval('Region', 'gridLimit')
def process_timeseries(config_file):
"""
Process a set of timeseries files to include the multiplier values.
The combined multiplier values are stored in a shape file as fields,
and records are keyed by the same code that is used to select
stations for sampling.
:param str config_file: Path to a configuration file.
"""
config = ConfigParser()
config.read(config_file)
stnFile = config.get('Timeseries', 'StationFile')
key_name = config.get('Timeseries', 'StationID')
tsPath = pjoin(config.get('Output', 'Path'),
'process', 'timeseries')
log.info("Loading stations from %s"%stnFile)
log.info("Timeseries data will be written into %s"%tsPath)
directions = ['n','ne','e','se','s','sw','w','nw']
sf = shapefile.Reader(stnFile)
field_names = [sf.fields[i][0] for i in range(1,len(sf.fields))]
key_index = field_names.index(key_name)
records = sf.records()
indexes = []
for d in directions:
fieldname = 'm4_%s' % d
indexes.append(field_names.index(fieldname))
for record in records:
stnId = record[key_index]
tsFile = pjoin(tsPath, 'ts.{0}.csv'.format(stnId))
if os.path.isfile(tsFile):
# Load multipliers for this location:
m = [float(record[i]) for i in indexes]
tsmultiply(tsFile, tuple(m))
else:
log.debug("No timeseries file for {0}".format(stnId))
pass
def main(configFile):
from Utilities.loadData import loadTrackFile
from Utilities.config import ConfigParser
from os.path import join as pjoin, normpath, dirname
baseDir = normpath(pjoin(dirname(__file__), '..'))
inputPath = pjoin(baseDir, 'input')
config = ConfigParser()
config.read(configFile)
inputFile = config.get('DataProcess', 'InputFile')
source = config.get('DataProcess', 'Source')
gridLimit = config.geteval('Region', 'gridLimit')
xx = np.arange(gridLimit['xMin'], gridLimit['xMax'] + .1, 0.1)
yy = np.arange(gridLimit['yMin'], gridLimit['yMax'] + .1, 0.1)
xgrid, ygrid = np.meshgrid(xx, yy)
if len(dirname(inputFile)) == 0:
inputFile = pjoin(inputPath, inputFile)
try:
tracks = loadTrackFile(configFile, inputFile, source)
except (TypeError, IOError, ValueError):
log.critical("Cannot load historical track file: {0}".format(inputFile))
raise
title = source
outputPath = config.get('Output', 'Path')
outputPath = pjoin(outputPath, 'plots','stats')
outputFile = pjoin(outputPath, 'tctracks.png')
map_kwargs = dict(llcrnrlon=xgrid.min(),
llcrnrlat=ygrid.min(),
urcrnrlon=xgrid.max(),
urcrnrlat=ygrid.max(),
projection='merc',
resolution='i')
figure = TrackMapFigure()
figure.add(tracks, xgrid, ygrid, title, map_kwargs)
figure.plot()
saveFigure(figure, outputFile)
def main(configFile):
from Utilities.loadData import loadTrackFile
from Utilities.config import ConfigParser
from os.path import join as pjoin, normpath, dirname
baseDir = normpath(pjoin(dirname(__file__), '..'))
inputPath = pjoin(baseDir, 'input')
config = ConfigParser()
config.read(configFile)
inputFile = config.get('DataProcess', 'InputFile')
source = config.get('DataProcess', 'Source')
gridLimit = config.geteval('Region', 'gridLimit')
xx = np.arange(gridLimit['xMin'], gridLimit['xMax'] + .1, 0.1)
yy = np.arange(gridLimit['yMin'], gridLimit['yMax'] + .1, 0.1)
xgrid, ygrid = np.meshgrid(xx, yy)
if len(dirname(inputFile)) == 0:
inputFile = pjoin(inputPath, inputFile)
try:
tracks = loadTrackFile(configFile, inputFile, source)
except (TypeError, IOError, ValueError):
log.critical("Cannot load historical track file: {0}".format(inputFile))
raise
title = source
outputPath = config.get('Output', 'Path')
outputPath = pjoin(outputPath, 'plots', 'stats')
outputFile = pjoin(outputPath, 'tctracks.png')
map_kwargs = dict(llcrnrlon=xgrid.min(),
llcrnrlat=ygrid.min(),
urcrnrlon=xgrid.max(),
urcrnrlat=ygrid.max(),
projection='merc',
resolution='i')
figure = TrackMapFigure()
figure.add(tracks, xgrid, ygrid, title, map_kwargs)
figure.plot()
saveFigure(figure, outputFile)
def colReadCSV(configFile, dataFile, source):
"""
Loads a csv file containing 'column' data into a record (numpy)
array with columns labelled by 'fields'. There must be a section in
the ``configFile`` named ``source`` that sets out the format of the
data file.
:param str configFile: Path to a configuration file that holds details
of the input data.
:param str dataFile: Path to the input file to load.
:param str source: Name of the source format. There must be a
corresponding section in the ``configFile``.
:returns: A :class:`numpy.ndarray` that contains the input data.
"""
config = ConfigParser()
config.read(configFile)
delimiter = config.get(source, 'FieldDelimiter')
numHeadingLines = config.getint(source, 'NumberOfHeadingLines')
cols = config.get(source, 'Columns').split(delimiter)
usecols = [i for i, c in enumerate(cols) if c != 'skip']
try:
data = np.genfromtxt(dataFile,
dtype=None,
delimiter=delimiter,
usecols=usecols,
comments=None,
skip_header=numHeadingLines,
autostrip=True)
except IOError:
log.exception("File not found: {0}".format(dataFile))
raise IOError("File not found: {0}".format(dataFile))
except TypeError:
log.exception(("{0} is not a string, filehandle "
"or generator.").format(dataFile))
raise TypeError(("{0} is not a string, filehandle "
"or generator.").format(dataFile))
data.dtype.names = [c for c in cols if c != 'skip']
return data
def copyTranslateMultipliers(configFile, type_mapping, output_path):
'''
Copy wind multipliers from directory specified in the configuration file, to an
output directory.
Once the files have been copied, they are translated into Geotiffs
:param str configFile: Path to configuration file
:param str type_mapping: dict containing the three wind multiplier inputs
'''
config = ConfigParser()
config.read(configFile)
# Check for wind multiplier file path in config file
if config.has_option('Input', 'Multipliers'):
WMPath = config.get('Input', 'Multipliers')
log.info('Using multiplier files from %s', WMPath)
else:
log.info(
'Using default multiplier files from /g/data/fj6/multipliers/')
WMPath = '/g/data/fj6/multipliers/'
tiles = config.get('Input', 'Tiles')
tiles = [item.strip() for item in tiles.split(',')]
log.info('Multipliers will be written out to %s', output_path)
checkOutputFolders(output_path)
for tile in tiles:
for wm in type_mapping:
var = type_mapping[wm]
pathn = WMPath + wm + '/'
log.debug('Copying files from %s', pathn)
log.debug('Beginning to translate tiles into Geotiff')
for file in glob.glob(pathn + tile + '*'):
file_break = file.split('/')
output = file_break[-1]
output_name = wm + '/' + output
copyfile(file, output_path + output_name)
os.system(
'gdal_translate -of GTiff NETCDF:{0}{1}/{2}:{3} {4}{5}.tif'
.format(output_path, wm, output, var, output_path,
output_name[:-3])) # -3 to drop '.nc'
log.debug('%s translated to Geotif', output_name[:-3])
def run(configFile, callback=None):
"""
Run the hazard calculations.
This will attempt to run the calculation in parallel by tiling the
domain, but also provides a sane fallback mechanism to execute
in serial.
:param configFile: str
"""
log.info("Loading hazard calculation settings")
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output', 'Path')
inputPath = pjoin(outputPath, 'windfield')
gridLimit = config.geteval('Region', 'gridLimit')
numsimulations = config.getint('TrackGenerator', 'NumSimulations')
yrsPerSim = config.getint('TrackGenerator', 'YearsPerSimulation')
minRecords = config.getint('Hazard', 'MinimumRecords')
calculate_confidence = config.getboolean('Hazard', 'CalculateCI')
wf_lon, wf_lat = setDomain(inputPath)
global pp
pp = attemptParallel()
log.info("Running hazard calculations")
TG = TileGrid(gridLimit, wf_lon, wf_lat)
tiles = getTiles(TG)
#def progress(i):
# callback(i, len(tiles))
pp.barrier()
hc = HazardCalculator(configFile, TG,
numsimulations,
minRecords,
yrsPerSim,
calculate_confidence)
hc.dumpHazardFromTiles(tiles)
pp.barrier()
hc.saveHazard()
log.info("Completed hazard calculation")
def colReadCSV(configFile, dataFile, source):
"""
Loads a csv file containing 'column' data into a record
(numpy) array with columns labelled by 'fields'.
"""
config = ConfigParser()
config.read(configFile)
delimiter = config.get(source, 'FieldDelimiter')
numHeadingLines = config.getint(source, 'NumberOfHeadingLines')
cols = config.get(source, 'Columns').split(delimiter)
usecols = [i for i,c in enumerate(cols) if c != 'skip']
data = np.genfromtxt(dataFile, dtype=None, delimiter=delimiter,
usecols=usecols, comments=None, skip_header=numHeadingLines,
autostrip=True)
data.dtype.names = [c for c in cols if c != 'skip']
return data
def main(configFile):
"""
Main function to execute the :mod:`wind`.
:param str configFile: Path to configuration file.
"""
config = ConfigParser()
config.read(configFile)
doOutputDirectoryCreation(configFile)
trackFile = config.get('DataProcess', 'InputFile')
source = config.get('DataProcess', 'Source')
delta = 1/12.
outputPath = pjoin(config.get('Output','Path'), 'tracks')
outputTrackFile = pjoin(outputPath, "tracks.interp.nc")
# This will save interpolated track data in TCRM format:
interpTrack = interpolateTracks.parseTracks(configFile, trackFile,
source, delta,
outputTrackFile,
interpolation_type='akima')
showProgressBar = config.get('Logging', 'ProgressBar')
pbar = ProgressBar('Calculating wind fields: ', showProgressBar)
def status(done, total):
pbar.update(float(done)/total)
import wind
wind.run(configFile, status)
import impact
impact.run_optional(config)
if config.getboolean('WindfieldInterface', 'PlotOutput'):
doWindfieldPlotting(configFile)
if config.getboolean('Timeseries', 'Extract'):
doTimeseriesPlotting(configFile)
def doDataDownload(configFile):
"""
Check and download the data files listed in the configuration file.
Datasets are listed in the `Input` section of the configuration
file, with the option `Datasets`. There must also be a corresponding
section in the configuration file that inlcudes the url, path where
the dataset will be stored and the filename that will be stored, e.g.::
[Input]
Datasets=IBTRACS
[IBTRACS]
URL=ftp://eclipse.ncdc.noaa.gov/pub/ibtracs/v03r05/wmo/csv/Allstorms.ibtracs_wmo.v03r05.csv.gz
filename=Allstorms.ibtracs_wmo.v03r05.csv
path=input
This will attempt to download the gzipped csv file from the given URL
and save it to the given filename, in the 'input' folder under the current
directory. Gzipped files are automatically unzipped.
:param str configFile: Name of configuration file.
:raises IOError: If the data cannot be downloaded.
"""
log.info('Checking availability of input data sets')
config = ConfigParser()
config.read(configFile)
showProgressBar = config.get('Logging', 'ProgressBar')
datasets.loadDatasets(configFile)
for dataset in datasets.DATASETS:
if not dataset.isDownloaded():
log.info('Input file %s is not available', dataset.filename)
try:
log.info('Attempting to download %s', dataset.filename)
pbar = ProgressBar('Downloading file %s: ' % dataset.filename,
showProgressBar)
def status(fn, done, size):
pbar.update(float(done)/size)
dataset.download(status)
log.info('Download successful')
except IOError:
log.error('Unable to download %s. Maybe a proxy problem?',
dataset.filename)
sys.exit(1)
def __init__(self, configFile, progressbar=None):
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output', 'Path')
try:
self.localityID = config.get('Region', 'LocalityID')
except Exception:
self.localityID = -999999
self.inputFile = pjoin(outputPath, 'hazard', 'hazard.nc')
self.plotPath = pjoin(outputPath, 'plots', 'hazard')
self.plotUnits = PlotUnits(config.get('Hazard', 'PlotSpeedUnits'))
self.ciBounds = config.getboolean('Hazard', 'CalculateCI')
self.fit = config.get('Hazard', 'ExtremeValueDistribution')
self.progressbar = progressbar
self.db = database.HazardDatabase(configFile)
def run(configFile):
"""
Run database update
:param str configFile: path to a configuration file.
"""
log.info('Running database update')
config = ConfigParser()
config.read(configFile)
outputPath = config.get('Output', 'Path')
location_db = pjoin(outputPath, 'locations.db')
if not os.path.exists(location_db):
location_file = config.get('Input', 'LocationFile')
buildLocationDatabase(location_db, location_file)
global MPI, comm
MPI = attemptParallel()
comm = MPI.COMM_WORLD
db = HazardDatabase(configFile)
db.createDatabase()
db.setLocations()
comm.barrier()
db.processEvents()
comm.barrier()
db.processHazard()
comm.barrier()
db.processTracks()
comm.barrier()
#db.close()
comm.barrier()
log.info("Created and populated database")
log.info("Finished running database creation")
def historic(self):
"""Load historic data and calculate histogram"""
log.info("Processing historical pressure distributions")
config = ConfigParser()
config.read(self.configFile)
inputFile = config.get('DataProcess', 'InputFile')
source = config.get('DataProcess', 'Source')
if len(os.path.dirname(inputFile)) == 0:
inputFile = pjoin(self.inputPath, inputFile)
try:
tracks = loadTrackFile(self.configFile, inputFile, source)
except (TypeError, IOError, ValueError):
log.critical(
"Cannot load historical track file: {0}".format(inputFile))
raise
else:
self.histMean, self.histMin, \
self.histMax, self.histMed = self.calculate(tracks)
self.histMinCPDist, self.histMinCP = self.calcMinPressure(tracks)