Fix CSV id normalization

1 files changed, 21 insertions, 8 deletions

diff --git a/data.py b/data.py
index 79b38c7..92efefc 100644
--- a/data.py
+++ b/data.py
@@ -13,22 +13,35 @@ if socket.gethostname() == "adeb.laptop":
 else:
     DATA_PATH="/data/lisatmp3/auvolat/taxikaggle"
 
-client_ids = {int(x): y+1 for y, x in enumerate(open(DATA_PATH+"/client_ids.txt"))}
-
-def get_client_id(n):
-    if n in client_ids:
-        return client_ids[n]
-    else:
-        return 0
+H5DATA_PATH = '/data/lisatmp3/simonet/taxi/data.hdf5'
 
 porto_center = numpy.array([41.1573, -8.61612], dtype=theano.config.floatX)
 data_std = numpy.sqrt(numpy.array([0.00549598, 0.00333233], dtype=theano.config.floatX))
 
-n_clients = 57124 #57105
+n_clients = 57124
+n_train_clients = 57105
 n_stands = 63
 
 dataset_size = 1710670
 
+# ---- Read client IDs and create reverse dictionnary
+
+def make_client_ids():
+    f = h5py.File(H5DATA_PATH, "r")
+    l = f['uniq_origin_call']
+    r = {}
+    for i in range(l.shape[0]):
+        r[l[i]] = i
+    return r
+
+client_ids = make_client_ids()
+
+def get_client_id(n):
+    if n in client_ids:
+        return client_ids[n]
+    else:
+        return 0
+
 class CallType(Enum):
     CENTRAL = 0
     STAND = 1