taxi

Winning entry to the Kaggle taxi competition
git clone https://esimon.eu/repos/taxi.git
Log | Files | Refs | README

commit e51b6fdf8b23dc1bb6331c34708a599f55b568f4
parent ff0d3a7e7e4d076d95e8aff0dbf2f1ea18473602
Author: Étienne Simon <esimon@esimon.eu>
Date:   Wed, 29 Apr 2015 11:17:41 -0400

Fix hdf5 polyline conversion

Diffstat:
Mconvert_data.py | 6+++---
1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/convert_data.py b/convert_data.py @@ -74,9 +74,9 @@ def read_taxis(input_directory, h5file, dataset, prefix): timestamp[id] = int(line[5]) day_type[id] = ord(line[6][0]) - ord('A') missing_data[id] = line[7][0] == 'T' - polyline = numpy.array(ast.literal_eval(line[8]), dtype=theano.config.floatX) - latitude[id] = polyline[:, 1] if polyline.ndim==2 else polyline - longitude[id] = polyline[:, 0] if polyline.ndim==2 else polyline + polyline = ast.literal_eval(line[8]) + latitude[id] = numpy.array([point[1] for point in polyline], dtype=theano.config.floatX) + longitude[id] = numpy.array([point[0] for point in polyline], dtype=theano.config.floatX) id+=1 splits = {} print >> sys.stderr, 'read %s: writing' % dataset