From 222971123f2741ee3689092fb4396dac83a13338 Mon Sep 17 00:00:00 2001
From: Alex Auvolat <alex.auvolat@ens.fr>
Date: Thu, 21 May 2015 17:05:07 -0400
Subject: Implement cut-based iteration scheme (SLOW!!!)

---
 data/cut.py             | 31 ++++++++++++++++++++++++++++++
 data/make_time_index.py | 50 +++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 81 insertions(+)
 create mode 100644 data/cut.py
 create mode 100755 data/make_time_index.py

(limited to 'data')

diff --git a/data/cut.py b/data/cut.py
new file mode 100644
index 0000000..1253434
--- /dev/null
+++ b/data/cut.py
@@ -0,0 +1,31 @@
+from fuel.schemes import IterationScheme
+import sqlite3
+import random
+import os
+from picklable_itertools import iter_
+
+import data
+
+first_time = 1372636853
+last_time = 1404172787
+
+
+class TaxiTimeCutScheme(IterationScheme):
+    def __init__(self, dbfile=None, use_cuts=None):
+        self.dbfile = os.path.join(data.path, 'time_index.db') if dbfile == None else dbfile
+        self.use_cuts = use_cuts
+
+    def get_request_iterator(self):
+        cuts = self.use_cuts
+        if cuts == None:
+            cuts = [random.randrange(first_time, last_time) for _ in range(100)]
+
+        l = []
+        with sqlite3.connect(self.dbfile) as db:
+            c = db.cursor()
+            for cut in cuts:
+                l = l + [i for (i,) in
+                    c.execute('SELECT trip FROM trip_times WHERE begin <= ? AND end >= ?', (cut, cut))]
+                
+        return iter_(l)
+
diff --git a/data/make_time_index.py b/data/make_time_index.py
new file mode 100755
index 0000000..c51d075
--- /dev/null
+++ b/data/make_time_index.py
@@ -0,0 +1,50 @@
+#!/usr/bin/env python
+# Make a valid dataset by cutting the training set at specified timestamps
+
+import os
+import sys
+import importlib
+
+import h5py
+import numpy
+
+import data
+from data.hdf5 import taxi_it
+
+import sqlite3
+
+def make_valid(outpath):
+    times = []
+    for i, line in enumerate(taxi_it('train')):
+        time = line['timestamp']
+        latitude = line['latitude']
+
+        if len(latitude) == 0:
+            continue
+
+        duration = 15 * (len(latitude) - 1)
+
+        times.append((i, int(time), int(time + duration)))
+        if i % 1000 == 0:
+            print times[-1]
+
+
+    with sqlite3.connect(outpath) as timedb:
+        c = timedb.cursor()
+        c.execute('''
+                CREATE TABLE trip_times
+                    (trip INTEGER, begin INTEGER, end INTEGER)
+        ''')
+        print "Adding data..."
+        c.executemany('INSERT INTO trip_times(trip, begin, end) VALUES(?, ?, ?)', times)
+        timedb.commit()
+        print "Creating index..."
+        c.execute('''CREATE INDEX trip_time_index ON trip_times (begin, end)''')
+
+
+if __name__ == '__main__':
+    if len(sys.argv) < 1 or len(sys.argv) > 2:
+        print >> sys.stderr, 'Usage: %s [outfile]' % sys.argv[0]
+        sys.exit(1)
+    outpath = os.path.join(data.path, 'time_index.db') if len(sys.argv) < 2 else sys.argv[1]
+    make_valid(outpath)
-- 
cgit v1.2.3