taxi

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

commit 61e0d47b6c6a570feebb43d474138020b13495aa
parent e51b6fdf8b23dc1bb6331c34708a599f55b568f4
Author: Alex Auvolat <alex.auvolat@ens.fr>
Date:   Wed, 29 Apr 2015 15:14:57 -0400

Add script for generation of map of destination points only.

Diffstat:
M.gitignore | 3+++
Adata_analysis/destmaps.py | 32++++++++++++++++++++++++++++++++
2 files changed, 35 insertions(+), 0 deletions(-)

diff --git a/.gitignore b/.gitignore @@ -60,3 +60,6 @@ target/ # Vim tmpfiles *.swp + +# Random datafiles +*.csv diff --git a/data_analysis/destmaps.py b/data_analysis/destmaps.py @@ -0,0 +1,32 @@ +import matplotlib.pyplot as plt +import numpy +import cPickle +import scipy.misc + +print "Loading data..." +with open("train_normal.pkl") as f: normal = cPickle.load(f) + +print "Extracting x and y" +# xes = [c[0] for l in normal for c in l[-1]] +# yes = [c[1] for l in normal for c in l[-1]] +xes = [l[-1][-1][0] for l in normal if len(l[-1]) > 0] +yes = [l[-1][-1][1] for l in normal if len(l[-1]) > 0] + +xrg = [-8.75, -8.55] +yrg = [41.05, 41.25] + +print "Doing 1d x histogram" +plt.clf(); plt.hist(xes, bins=1000, range=xrg); plt.savefig("xhist_dest.pdf") +print "Doing 1d y histogram" +plt.clf(); plt.hist(yes, bins=1000, range=yrg); plt.savefig("yhist_dest.pdf") + +print "Doing 2d histogram" +hist, xx, yy = numpy.histogram2d(xes, yes, bins=2000, range=[xrg, yrg]) + +# import ipdb; ipdb.set_trace() + +print "Imshow" +plt.clf(); plt.imshow(numpy.log(hist)); plt.savefig("xyhmap_dest.png", dpi=600) + +print "Imsave" +scipy.misc.imsave("xymap_dest_2.png", numpy.log(hist + 1))