def run(params, logger):
"""."""
logger.debug("Loading parameters.")
prod_mode = params.get('prod_mode')
calc_traj = params.get('calc_traj')
calc_state = params.get('calc_state')
calc_perf = params.get('calc_perf')
prod_traj_year_initial = int(params.get('prod_traj_year_initial'))
prod_traj_year_final = int(params.get('prod_traj_year_final'))
prod_perf_year_initial = int(params.get('prod_perf_year_initial'))
prod_perf_year_final = int(params.get('prod_perf_year_final'))
prod_state_year_bl_start = int(params.get('prod_state_year_bl_start'))
prod_state_year_bl_end = int(params.get('prod_state_year_bl_end'))
prod_state_year_tg_start = int(params.get('prod_state_year_tg_start'))
prod_state_year_tg_end = int(params.get('prod_state_year_tg_end'))
geojsons = json.loads(params.get('geojsons'))
crs = params.get('crs')
prod_traj_method = params.get('trajectory_method')
ndvi_gee_dataset = params.get('ndvi_gee_dataset')
climate_gee_dataset = params.get('climate_gee_dataset')
# crs = ee.Image("users/geflanddegradation/toolbox_datasets/ndvi_modis_2001_2019").projection().crs()
# Check the ENV. Are we running this locally or in prod?
if params.get('ENV') == 'dev':
EXECUTION_ID = str(random.randint(1000000, 99999999))
else:
EXECUTION_ID = params.get('EXECUTION_ID', None)
logger.debug("Running productivity indicators.")
proj = ee.Image(ndvi_gee_dataset).projection()
if prod_mode == 'Trends.Earth productivity':
outs = []
for geojson in geojsons:
this_out = None
if calc_traj:
traj = productivity_trajectory(geojsons[0]['coordinates'],
int(prod_traj_year_initial),
int(prod_traj_year_final),
prod_traj_method,
ndvi_gee_dataset,
climate_gee_dataset, logger)
if not this_out:
this_out = traj
if calc_perf:
perf = productivity_performance(geojsons[0]['coordinates'],
prod_perf_year_initial,
prod_perf_year_final,
ndvi_gee_dataset, geojson,
EXECUTION_ID, logger)
if not this_out:
this_out = perf
else:
this_out.merge(perf)
if calc_state:
state = productivity_state(
geojsons[0]['coordinates'], prod_state_year_bl_start,
prod_state_year_bl_end, prod_state_year_tg_start,
prod_state_year_tg_end, ndvi_gee_dataset, EXECUTION_ID,
logger)
if not this_out:
this_out = state
else:
this_out.merge(state)
outs.append(
this_out.export([geojson], 'productivity', crs, logger,
EXECUTION_ID, proj))
# First need to deserialize the data that was prepared for output from
# the productivity functions, so that new urls can be appended
schema = CloudResultsSchema()
logger.debug("Deserializing")
final_output = schema.load(outs[0])
for o in outs[1:]:
this_out = schema.load(o)
final_output.urls.extend(this_out.urls)
logger.debug("Serializing")
# Now serialize the output again and return it
return schema.dump(final_output)
elif prod_mode == 'JRC LPD':
out = download(
'users/geflanddegradation/toolbox_datasets/lpd_300m_longlat',
'Land Productivity Dynamics (LPD)', 'one time', None, None,
EXECUTION_ID, logger)
return out.export(geojsons, 'productivity', crs, logger, EXECUTION_ID,
proj)
else:
raise Exception(
'Unknown productivity mode "{}" chosen'.format(prod_mode))
def run(params, logger):
"""."""
logger.debug("Loading parameters.")
prod_mode = params.get('prod_mode')
prod_traj_year_initial = params.get('prod_traj_year_initial')
prod_traj_year_final = params.get('prod_traj_year_final')
prod_perf_year_initial = params.get('prod_perf_year_initial')
prod_perf_year_final = params.get('prod_perf_year_final')
prod_state_year_bl_start = params.get('prod_state_year_bl_start')
prod_state_year_bl_end = params.get('prod_state_year_bl_end')
prod_state_year_tg_start = params.get('prod_state_year_tg_start')
prod_state_year_tg_end = params.get('prod_state_year_tg_end')
lc_year_initial = params.get('lc_year_initial')
lc_year_final = params.get('lc_year_final')
soc_year_initial = params.get('soc_year_initial')
soc_year_final = params.get('soc_year_final')
geojsons = json.loads(params.get('geojsons'))
crs = params.get('crs')
prod_traj_method = params.get('prod_traj_method')
ndvi_gee_dataset = params.get('ndvi_gee_dataset')
climate_gee_dataset = params.get('climate_gee_dataset')
fl = params.get('fl')
trans_matrix = params.get('trans_matrix')
remap_matrix = params.get('remap_matrix')
if len(trans_matrix) != 49:
raise GEEIOError("Transition matrix must be a list with 49 entries")
if len(remap_matrix) != 2 or len(remap_matrix[0]) != 37 or len(remap_matrix[1]) != 37:
raise GEEIOError("Transition matrix must be a list of two lists with 37 entries each")
# Check the ENV. Are we running this locally or in prod?
if params.get('ENV') == 'dev':
EXECUTION_ID = str(random.randint(1000000, 99999999))
else:
EXECUTION_ID = params.get('EXECUTION_ID', None)
proj = ee.Image(ndvi_gee_dataset).projection()
logger.debug("Running productivity indicators.")
if prod_mode == 'Trends.Earth productivity':
outs = []
for geojson in geojsons:
# Need to loop over the geojsons, since performance takes in a
# geojson.
# TODO: pass performance a second geojson defining the entire extent of
# all input geojsons so that the performance is calculated the same
# over all input areas.
out = productivity_trajectory(prod_traj_year_initial,
prod_traj_year_final, prod_traj_method,
ndvi_gee_dataset, climate_gee_dataset,
logger)
prod_perf = productivity_performance(prod_perf_year_initial,
prod_perf_year_final,
ndvi_gee_dataset, geojson,
EXECUTION_ID, logger)
out.merge(prod_perf)
prod_state = productivity_state(prod_state_year_bl_start,
prod_state_year_bl_end,
prod_state_year_tg_start,
prod_state_year_tg_end,
ndvi_gee_dataset, EXECUTION_ID, logger)
out.merge(prod_state)
logger.debug("Running land cover indicator.")
lc = land_cover(lc_year_initial, lc_year_final, trans_matrix, remap_matrix,
EXECUTION_ID, logger)
lc.selectBands(['Land cover (degradation)',
'Land cover (7 class)'])
out.merge(lc)
logger.debug("Running soil organic carbon indicator.")
soc_out = soc(soc_year_initial, soc_year_final, fl, remap_matrix, False,
EXECUTION_ID, logger)
soc_out.selectBands(['Soil organic carbon (degradation)',
'Soil organic carbon'])
out.merge(soc_out)
out.setVisible(['Soil organic carbon (degradation)',
'Land cover (degradation)',
'Productivity trajectory (significance)',
'Productivity state (degradation)',
'Productivity performance (degradation)',
'Land Productivity Dynamics (LPD)'])
outs.append(out.export([geojson], 'sdg_sub_indicators', crs, logger,
EXECUTION_ID, proj))
# First need to deserialize the data that was prepared for output from
# the productivity functions, so that new urls can be appended
schema = CloudResultsSchema()
logger.debug("Deserializing")
final_prod = schema.load(outs[0])
for o in outs[1:]:
this_out = schema.load(o)
final_prod.urls.extend(this_out.urls)
logger.debug("Serializing")
# Now serialize the output again so the remaining layers can be added
# to it
return schema.dump(final_prod)
elif prod_mode == 'JRC LPD':
out = download('users/geflanddegradation/toolbox_datasets/lpd_300m_longlat',
'Land Productivity Dynamics (LPD)', 'one time',
None, None, EXECUTION_ID, logger)
# Save as int16 to be compatible with other data
out.image = out.image.int16()
logger.debug("Running land cover indicator.")
lc = land_cover(lc_year_initial, lc_year_final, trans_matrix, remap_matrix,
EXECUTION_ID, logger)
lc.selectBands(['Land cover (degradation)',
'Land cover (7 class)'])
out.merge(lc)
logger.debug("Running soil organic carbon indicator.")
soc_out = soc(soc_year_initial, soc_year_final, fl, remap_matrix, False,
EXECUTION_ID, logger)
soc_out.selectBands(['Soil organic carbon (degradation)',
'Soil organic carbon'])
out.merge(soc_out)
out.setVisible(['Soil organic carbon (degradation)',
'Land cover (degradation)',
'Productivity trajectory (significance)',
'Productivity state (degradation)',
'Productivity performance (degradation)',
'Land Productivity Dynamics (LPD)'])
return out.export(geojsons, 'sdg_sub_indicators', crs, logger,
EXECUTION_ID, proj)
else:
raise Exception('Unknown productivity mode "{}" chosen'.format(prod_mode))