est = model_data['model']
params = model_data['params']
num_folds = model_data['num_folds']
fuel_types = model_data['fuel_types']
print("Loaded trained generation estimation model from {0}.".format(
args.model_filename))
for k, v in params.iteritems():
print(" - {0}: {1}".format(k, v))
print(" - num_folds: {0}".format(num_folds))
print("Fuel types: {0}".format(fuel_types))
# load powerplant database
if args.powerplant_database.endswith('.csv'):
plants = pw.read_csv_file_to_dict(args.powerplant_database)
else:
plants = pw.load_database(args.powerplant_database)
print("Loaded {0} plants from file {1}.".format(len(plants),
args.powerplant_database))
# set up arrays
fuel_type_list = []
capacity_by_country_by_fuel = {
} # will hold total capacity by country and fuel
feature_name_list = [
'fuel_type', 'capacity_mw', 'commissioning_year', 'fuel_avg_cf',
'cap_sh_country', 'cap_sh_country_fuel'
]
# read data from plant database
count_full_data = 0
powerwatch_datadump = {}
# make plant condcordance dictionary
plant_concordance = pw.make_plant_concordance()
print("Loaded concordance file with {0} entries.".format(len(plant_concordance)))
carma_id_used = [] # Record matched carma_ids
# STEP 0: Read in source databases.
# Identify countries with automated data from .has_api flag.
print("Loading source databases...")
country_databases = {}
for country_name,country in country_dictionary.iteritems():
if country.has_api == 1:
country_code = country.iso_code
database_filename = COUNTRY_DATABASE_FILE.replace("COUNTRY", country_code)
country_databases[country_name] = pw.load_database(database_filename)
print("Loaded {0} plants from {1} database.".format(len(country_databases[country_name]),country_name))
# Load multi-country databases.
wri_database = pw.load_database(WRI_DATABASE_FILE)
print("Loaded {0} plants from WRI database.".format(len(wri_database)))
geo_database = pw.load_database(GEO_DATABASE_FILE)
print("Loaded {0} plants from GEO database.".format(len(geo_database)))
carma_database = pw.load_database(CARMA_DATABASE_FILE)
print("Loaded {0} plants from CARMA database.".format(len(carma_database)))
sourcewatch_database = pw.load_database(SOURCEWATCH_DATABASE_FILE)
print("Loaded {0} plants from SourceWatch database.".format(len(sourcewatch_database)))
# Track counts using a dict with keys corresponding to each data source
db_sources = country_databases.keys()
db_sources.extend(["WRI","GEO","SourceWatch","WRI with GEO lat/long data","WRI with CARMA lat/long data"])
# True if specified --download, otherwise False
FILES = {RAW_FILE_NAME_REPD: URL_REPD, RAW_FILE_NAME_DUKES: URL_DUKES}
DOWNLOAD_FILES = pw.download(
u"UK Renewable Energy Planning Database and DUKES", FILES)
# set up fuel type thesaurus
fuel_thesaurus = pw.make_fuel_thesaurus()
# set up country name thesaurus
country_thesaurus = pw.make_country_names_thesaurus()
# create dictionary for power plant objects
plants_dictionary = {}
# load GEO and CARMA for matching coordinates
geo_database = pw.load_database(GEO_DATABASE_FILE)
print("Loaded {0} plants from GEO database.".format(len(geo_database)))
carma_database = pw.load_database(CARMA_DATABASE_FILE)
print("Loaded {0} plants from CARMA database.".format(len(carma_database)))
# read in plant matches file
with open(PLANT_MATCHES, "rbU") as f:
f.readline() # skip headers
csvreader = csv.reader(f)
plant_matches = {}
for row in csvreader:
dukes_name = str(row[0])
geo_id = pw.make_id(SAVE_CODE_GEO, int(row[1])) if row[1] else ""
carma_id = pw.make_id(SAVE_CODE_CARMA, int(row[2])) if row[2] else ""
repd_id = int(row[3]) if row[3] else ""
plant_matches[dukes_name] = {