def bayesian_adjust_file(prior_filename, data_file, output_base, res,
lower_left_coord, upper_right_coord):
"""Adjust the file given in data file using a Bayesian approach,
treating the given prior_filename as an image representing the prior,
and computing a new modern coordinate pair as the MAP of the posterior
resulting, all using grid approximation."""
lon_lim = lower_left_coord[0], upper_right_coord[0]
lat_lim = upper_right_coord[1], lower_left_coord[1]
prior = ImagePrior(prior_filename, lat_lim, lon_lim, res)
places = pd.read_csv(data_file, encoding='utf-8')
places.rename(columns={
'modern_lat': 'original_lat',
'modern_lon': 'original_lon'
},
inplace=True)
known = places[places.disposition == 'known']
known.is_copy = False
known.ix[:, 'modern_lat'] = known.ix[:, 'original_lat']
known.ix[:, 'modern_lon'] = known.ix[:, 'original_lon']
unknown = places[places.disposition != 'known']
unknown.is_copy = False
adjusted = unknown.apply(prior.bayesian_adjust, axis=1)
unknown = unknown.merge(adjusted, left_index=True, right_index=True)
kml_filename = os.path.join(PTOL_HOME, 'Data', output_base + '.kml')
csv_filename = os.path.join(PTOL_HOME, 'Data', output_base + '.csv')
common.write_kml_file(kml_filename, None, known, unknown)
common.write_csv_file(csv_filename, known, unknown)
def main(filename, model, places):
known, unknown = common.split_places(places)
knownx = known.loc[:, XCOLS]
knowny = known.loc[:, YCOLS]
model.fit(knownx, knowny)
unknownx = unknown.loc[:, XCOLS]
unknowny = model.predict(unknownx)
unknown.loc[:, YCOLS] = unknowny
title = ' '.join(os.path.basename(filename)[0:-4].split('_'))
common.write_kml_file(filename, None, known, unknown)
common.write_csv_file(filename[0:-4] + '.csv', known, unknown)
common.write_map_file(filename[0:-4] + '.pdf', known, unknown, 30, 24, 300, 'ptol_name', title)
common.write_map_file(filename[0:-4] + '.png', known, unknown, 30, 24, 300, 'ptol_name', title)
def main(filename, model, places):
known, unknown = common.split_places(places)
knownx = known.loc[:, XCOLS]
knowny = known.loc[:, YCOLS]
model.fit(knownx, knowny)
unknownx = unknown.loc[:, XCOLS]
unknowny = model.predict(unknownx)
unknown.loc[:, YCOLS] = unknowny
title = ' '.join(os.path.basename(filename)[0:-4].split('_'))
common.write_kml_file(filename, None, known, unknown)
common.write_csv_file(filename[0:-4] + '.csv', known, unknown)
common.write_map_file(filename[0:-4] + '.pdf', known, unknown, 30, 24, 300,
'ptol_name', title)
common.write_map_file(filename[0:-4] + '.png', known, unknown, 30, 24, 300,
'ptol_name', title)
def bayesian_adjust_file(prior_filename, data_file, output_base, res):
"""Adjust the file given in data file using a Bayesian approach,
treating the given prior_filename as an image representing the prior,
and computing a new modern coordinate pair as the MAP of the posterior
resulting, all using grid approximation."""
prior = ImagePrior(prior_filename, (35,5), (65,95), res)
places = pd.read_csv(data_file, encoding='cp1252')
places.rename(columns={
'modern_lat': 'original_lat',
'modern_lon': 'original_lon'}, inplace=True)
known = places[places.disposition == 'known']
known.is_copy = False
known.ix[:, 'modern_lat'] = known.ix[:, 'original_lat']
known.ix[:, 'modern_lon'] = known.ix[:, 'original_lon']
unknown = places[places.disposition == 'unknown']
unknown.is_copy = False
adjusted = unknown.apply(prior.bayesian_adjust, axis=1)
unknown = unknown.merge(adjusted, left_index=True, right_index=True)
kml_filename = os.path.join(PTOL_HOME, 'Data', output_base+'.kml')
csv_filename = os.path.join(PTOL_HOME, 'Data', output_base+'.csv')
common.write_kml_file(kml_filename, None, known, unknown)
common.write_csv_file(csv_filename, known, unknown)
import common
import extra
import constant
# Get Price Data from URL
price_data = common.get_data(constant.URL)['series'][0]['data']
# Format Date for each row
for i, date_and_price in enumerate(price_data):
price_data[i][0] = common.format_date(date_and_price[0])
# Write daily prices to data folder
common.write_csv_file(constant.DAILY_PRICE_FILE_PATH, constant.HEADING,
price_data)
#write monthly data to csv with normalization
common.write_csv_file(constant.MONTHLY_PRICE_FILE_PATH, constant.HEADING,
extra.normalize_to_monthly(price_data))
