taxi

commit 8d4621c30a4926a3393733175390bea3d3c138a0
parent f34ecb21c0ec362b560c2fe1c3afacc1e6dad998
Author: Alex Auvolat <alex.auvolat@ens.fr>
Date:   Fri, 24 Apr 2015 13:45:39 -0400

Add histogram producing script.

Diffstat:
A
alex/plots.py
 | 
29
+++++++++++++++++++++++++++++
A
data/test.csv
 | 
2
++
A
data/train.csv
 | 
2
++

3 files changed, 33 insertions(+), 0 deletions(-)
diff --git a/alex/plots.py b/alex/plots.py
@@ -0,0 +1,29 @@
+import matplotlib.pyplot as plt
+import numpy
+import cPickle
+import scipy
+
+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]]
+
+xrg = [-8.75, -8.55]
+yrg = [41.05, 41.25]
+
+print "Doing 1d histogram"
+#plt.clf(); plt.hist(xes, bins=1000, range=xrg); plt.savefig("xhist.pdf")
+#plt.clf(); plt.hist(yes, bins=1000, range=yrg); plt.savefig("yhist.pdf")
+
+print "Doing 2d histogram"
+#plt.clf(); plt.hist2d(xes, yes, bins=500, range=[xrg, yrg]); plt.savefig("xymap.pdf")
+
+hist, xx, yy = numpy.histogram2d(xes, yes, bins=2000, range=[xrg, yrg])
+
+import ipdb; ipdb.set_trace()
+
+plt.clf(); plt.imshow(numpy.log(hist)); plt.savefig("xyhmap.pdf")
+
+scipy.misc.imsave("xymap.png", numpy.log(hist))
diff --git a/data/test.csv b/data/test.csv
@@ -0,0 +1 @@
+/data/lisatmp3/auvolat/taxikaggle/test.csv+
\ No newline at end of file
diff --git a/data/train.csv b/data/train.csv
@@ -0,0 +1 @@
+/data/lisatmp3/auvolat/taxikaggle/train.csv+
\ No newline at end of file