return result
def parse_args():
"""
:returns: arguments
"""
import argparse
parser = argparse.ArgumentParser()
parser.add_argument('image_dir_path', help='Path to directory that contains images')
parser.add_argument('chunk_count', help='Number of chunks to split input paths into', type=int)
parser.add_argument('process_chunk', help='Chunk that will be processed', type=int)
parser.add_argument('output_name', help='Custom name added to the output')
return parser.parse_args()
if __name__ == '__main__':
args = parse_args()
image_dir_path = args.image_dir_path
chunk_count = args.chunk_count
process_chunk = args.process_chunk
output_name = args.output_name
debug_output_path = './output/detection_debug'
detections = detect_traffic_signs(image_dir_path, chunk_count=chunk_count, process_chunk=process_chunk, debug_output_path=debug_output_path)
detections_output_path = './detections_{}_chunk_{}_of_{}.pickle'.format(output_name, process_chunk, chunk_count)
util.pickle_save(detections_output_path, detections)
import glob
import numpy as np
from os.path import join, basename
import util
def get_pickle_files(dir_path):
return sorted(glob.glob(join(dir_path, '*.pickle')))
DATASET_NAME = '10_right'
DIR_PATH = join('./output/scores/', DATASET_NAME)
OUTPUT_PATH = join('./output/scores/merged', DATASET_NAME + '.pickle')
pickle_files = get_pickle_files(DIR_PATH)
result = {}
for pickle_file in pickle_files:
image_name = basename(pickle_file)[:-len('.pickle')]
scores = util.pickle_load(pickle_file)
result[image_name] = scores
#print(result[list(result.keys())[0]].shape)
util.pickle_save(OUTPUT_PATH, result)
#!/usr/bin/python3
# -*- coding: utf-8 -*-
from util import pickle_save
from database import ParkrunDB
#norwich_db = ParkrunDB("norwich", 440, 441)
#norwich_db.update()
#pickle_save(norwich_db, 'norwich_parkrun_2019.db')
colneylane_db = ParkrunDB("colneylane", 56, 57)
colneylane_db.update()
pickle_save(colneylane_db, 'colneylane_parkrun_2019.db')
def gettripIds(index=0):
tripIDs = []
# For each route,
for routeID, routeTrips in weekdayTrips.groupby('route_id'):
# Pick a trip
tripIDs.append(routeTrips.trip_id.values[index])
return tripIDs
if __name__ == "__main__":
calendar = read_csv('subwaydata/google_transit/calendar.txt')
routes = read_csv('subwaydata/google_transit/routes.txt')
trips = read_csv('subwaydata/google_transit/trips.txt')
times = read_csv('subwaydata/google_transit/stop_times.txt')
stops = read_csv('subwaydata/google_transit/stops.txt')
weekdayServiceIDs = filter(lambda x: x.endswith('WKD'),
calendar.service_id)
routeNameByID = {x['route_id']:
x['route_long_name'] for index, x in routes.iterrows()}
weekdayTrips = trips[trips.service_id.isin(weekdayServiceIDs)]
print(len(weekdayTrips))
weekdayTimes = times[times.trip_id.isin(weekdayTrips.trip_id.unique())]
tripIDs = gettripIds()
graph = make_graph(tripIDs)
import util
util.pickle_save(graph, 'subwaydata/NYCsubway_network_graph.pkl')
totalzipcodes = 0
missing = 0
for neighborhood in neighborhooddict.keys():
for zipcode in neighborhooddict[neighborhood]:
totalzipcodes += 1
try:
len(fulldata)
tmpdata = getData(zipcode)
if tmpdata is None:
missing += 1
continue
else:
fulldata = pd.concat([fulldata, tmpdata], axis=1)
filename = 'data/%s_%s.pkl' % (neighborhood, zipcode)
util.pickle_save(tmpdata, filename)
except NameError:
# hack to instantiate first DataFrame
fulldata = getData(zipcode)
filename = 'data/%s_%s.pkl' % (neighborhood, zipcode)
util.pickle_save(fulldata, filename)
print('Missing %d zipcodes of %d total' % (missing, totalzipcodes))
for zipcode in fulldata.keys():
plot_date(fulldata[zipcode].index,
fulldata[zipcode], alpha=0.7, fmt='.')
plot_date(fulldata.index, fulldata.mean(axis=1))