"""
Opendrift module
.. currentmodule:: opendrift
.. doctest::
>>> import opendrift
"""
import logging; logger = logging.getLogger(__name__)
import importlib
import numpy as np
from .version import __version__
# For automated access to available drift classes, e.g. for GUI
# Hardcoded for now
_available_models = \
['leeway.Leeway',
'openoil.OpenOil',
'larvalfish.LarvalFish',
'plastdrift.PlastDrift',
'shipdrift.ShipDrift',
'openberg_old.OpenBergOld']
[docs]
def get_model_names():
return [m.split('.')[-1] for m in _available_models]
[docs]
def get_model(model_name):
if model_name not in get_model_names():
raise ValueError('No drift model named %s' % model_name)
else:
for m in _available_models:
if m.split('.')[-1] == model_name:
module = importlib.import_module(
'opendrift.models.' + m.split('.')[0])
model = getattr(module, model_name)
return model
[docs]
def open(filename, times=None, elements=None, load_history=True):
'''Import netCDF output file as OpenDrift object of correct class'''
import os
import pydoc
from netCDF4 import Dataset
if not os.path.exists(filename):
logger.info('File does not exist, trying to retrieve from URL')
import urllib
try:
urllib.urlretrieve(filename, 'opendrift_tmp.nc')
filename = 'opendrift_tmp.nc'
except:
raise ValueError('%s does not exist' % filename)
n = Dataset(filename)
try:
module_name = n.opendrift_module
class_name = n.opendrift_class
except:
logger.warning(filename + ' does not contain global attributes '
'opendrift_module and opendrift_class, defaulting to OceanDrift')
module_name = 'oceandrift'
class_name = 'OceanDrift'
n.close()
if class_name == 'OpenOil3D':
class_name = 'OpenOil'
module_name = 'opendrift.models.openoil'
if class_name == 'OceanDrift3D':
class_name = 'OceanDrift'
module_name = 'opendrift.models.oceandrift'
cls = pydoc.locate(module_name + '.' + class_name)
if cls is None:
from opendrift.models import oceandrift
cls = oceandrift.OceanDrift
o = cls()
o.io_import_file(filename, times=times, elements=elements, load_history=load_history)
logger.info('Returning ' + str(type(o)) + ' object')
return o
[docs]
def open_xarray(filename, chunks={'trajectory': 50000, 'time': 1000}, elements=None):
'''Import netCDF output file as OpenDrift object of correct class'''
import os
import pydoc
import xarray as xr
if not os.path.exists(filename):
logger.info('File does not exist, trying to retrieve from URL')
import urllib
try:
urllib.urlretrieve(filename, 'opendrift_tmp.nc')
filename = 'opendrift_tmp.nc'
except:
raise ValueError('%s does not exist' % filename)
n = xr.open_dataset(filename)
try:
module_name = n.opendrift_module
class_name = n.opendrift_class
except:
raise ValueError(filename + ' does not contain '
'necessary global attributes '
'opendrift_module and opendrift_class')
n.close()
if class_name == 'OpenOil3D':
class_name = 'OpenOil'
module_name = 'opendrift.models.openoil'
if class_name == 'OceanDrift3D':
class_name = 'OceanDrift'
module_name = 'opendrift.models.oceandrift'
cls = pydoc.locate(module_name + '.' + class_name)
if cls is None:
from opendrift.models import oceandrift
cls = oceandrift.OceanDrift
o = cls()
o.io_import_file_xarray(filename, chunks=chunks, elements=elements)
logger.info('Returning ' + str(type(o)) + ' object')
return o
[docs]
def versions():
import multiprocessing
import platform
import scipy
import matplotlib
import netCDF4
import xarray
try:
import adios_db
adios_version = adios_db.__version__
except:
adios_version = ': Not installed'
try:
import copernicusmarine
copernicus_version = copernicusmarine.__version__
except:
copernicus_version = ': Not installed'
import sys
s = '\n------------------------------------------------------\n'
s += 'Software and hardware:\n'
s += ' OpenDrift version %s\n' % __version__
s += ' Platform: %s, %s\n' % (platform.system(), platform.release())
try:
from psutil import virtual_memory
ram = virtual_memory().total/(1024**3)
except:
ram = 'unknown'
s += ' %s GB memory\n' % ram
s += ' %s processors (%s)\n' % (multiprocessing.cpu_count(),
platform.processor())
s += ' NumPy version %s\n' % np.__version__
s += ' SciPy version %s\n' % scipy.__version__
s += ' Matplotlib version %s\n' % matplotlib.__version__
s += ' NetCDF4 version %s\n' % netCDF4.__version__
s += ' Xarray version %s\n' % xarray.__version__
s += ' ADIOS (adios_db) version %s\n' % adios_version
s += ' Copernicusmarine version %s\n' % copernicus_version
s += ' Python version %s\n' % sys.version.replace('\n', '')
s += '------------------------------------------------------\n'
return s
[docs]
def import_from_ladim(ladimfile, romsfile):
"""Import Ladim output file as OpenDrift simulation obejct"""
from models.oceandrift import OceanDrift
o = OceanDrift()
from netCDF4 import Dataset, date2num, num2date
if isinstance(romsfile, str):
from opendrift.readers import reader_ROMS_native
romsfile = reader_ROMS_native.Reader(romsfile)
l = Dataset(ladimfile, 'r')
pid = l.variables['pid'][:]
particle_count = l.variables['particle_count'][:]
end_index = np.cumsum(particle_count)
start_index = np.concatenate(([0], end_index[:-1]))
x = l.variables['X'][:]
y = l.variables['Y'][:]
lon, lat = romsfile.xy2lonlat(x, y)
time = num2date(l.variables['time'][:],
l.variables['time'].units)
history_dtype_fields = [
(name, o.ElementType.variables[name]['dtype'])
for name in o.ElementType.variables]
# Add environment variables
o.history_metadata = o.ElementType.variables.copy()
history_dtype = np.dtype(history_dtype_fields)
num_timesteps = len(time)
num_elements = len(l.dimensions['particle'])
o.history = np.ma.array(
np.zeros([num_elements, num_timesteps]),
dtype=history_dtype, mask=[True])
for n in range(num_timesteps):
start = start_index[n]
active = pid[start:start+particle_count[n]]
o.history['lon'][active, n] = \
lon[start:start+particle_count[n]]
o.history['lat'][active, n] = \
lat[start:start+particle_count[n]]
o.history['status'][active, n] = 0
o.status_categories = ['active', 'missing_data']
firstlast = np.ma.notmasked_edges(o.history['status'], axis=1)
index_of_last = firstlast[1][1]
o.history['status'][np.arange(len(index_of_last)),
index_of_last] = 1
kwargs = {}
for var in ['lon', 'lat', 'status']:
kwargs[var] = o.history[var][
np.arange(len(index_of_last)), index_of_last]
kwargs['ID'] = range(num_elements)
o.elements = o.ElementType(**kwargs)
o.elements_deactivated = o.ElementType()
o.remove_deactivated_elements()
# Import time steps from metadata
o.time_step = time[1] - time[0]
o.time_step_output = o.time_step
o.start_time = time[0]
o.time = time[-1]
o.steps_output = num_timesteps
return o