Added More Info

More information is added to make interested readers happier.
mmalekzadeh · Feb 17, 2020 · d950254 · d950254
1 parent 1b94304
commit d950254
Show file tree

Hide file tree

Showing 12 changed files with 1,723 additions and 3 deletions.
diff --git a/experiments/2_criteo_exp.ipynb b/experiments/2_criteo_exp.ipynb
@@ -69,6 +69,14 @@
     "from keras import backend as K"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Note:\n",
+    "`ads_f_name` is the address (directory+filename) of the created dataset that you chose in`create_datasets.ipynb`. If you do not use our default `train_1000000000_rows`, then you also need to set `ads_build` to `True` for the first time."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 3,

diff --git a/experiments/5_a_synthetic_figure_5.ipynb b/experiments/5_a_synthetic_figure_5.ipynb
diff --git a/experiments/Criteo/criteo_dataset/create_datasets.ipynb b/experiments/Criteo/criteo_dataset/create_datasets.ipynb
@@ -6,7 +6,9 @@
    "source": [
     "### Note:\n",
     "Before running this you need to download datasets from this address:\n",
-    "http://labs.criteo.com/2014/02/kaggle-display-advertising-challenge-dataset//"
+    "http://labs.criteo.com/2014/02/kaggle-display-advertising-challenge-dataset//\n",
+    "\n",
+    "Then, you need to choose your desired `nrows`. In our paper we use `nrows=1000000000`."
    ]
   },
   {
@@ -54,7 +56,9 @@
     "    \n",
     "    data.to_csv(data_type+\"_\"+str(nrows)+\"_rows.csv\", index=False)\n",
     "    print(data.shape)\n",
-    "nrows = 10000000\n",
+    "\n",
+    "##### Please choose your desired number of rows \n",
+    "nrows = 1000000000\n",
     "create_dataset(nrows, data_type=\"train\")"
    ]
   },
@@ -64,7 +68,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "pd.read_csv(\"train.csv\", nrows=nrows)"
+    "pd.read_csv(\"train_\"+str(nrows)+\"_rows.csv\", nrows=nrows)"
    ]
   }
  ],

diff --git a/experiments/SysML_Synthetic_drng_20_20_.npy b/experiments/SysML_Synthetic_drng_20_20_.npy
diff --git a/experiments/encoders_repo/kmeans_hasher_6_1_10.pkl b/experiments/encoders_repo/kmeans_hasher_6_1_10.pkl
diff --git a/experiments/encoders_repo/kmeans_hasher_7_1_10.pkl b/experiments/encoders_repo/kmeans_hasher_7_1_10.pkl
diff --git a/experiments/encoders_repo/kmeans_hasher_8_1_10.pkl b/experiments/encoders_repo/kmeans_hasher_8_1_10.pkl
diff --git a/experiments/encoders_repo/kmeans_hasher_9_1_10.pkl b/experiments/encoders_repo/kmeans_hasher_9_1_10.pkl
diff --git a/experiments/encoders_repo/re_indexer_hasher_6_1_10.npy b/experiments/encoders_repo/re_indexer_hasher_6_1_10.npy
diff --git a/experiments/encoders_repo/re_indexer_hasher_7_1_10.npy b/experiments/encoders_repo/re_indexer_hasher_7_1_10.npy
diff --git a/experiments/encoders_repo/re_indexer_hasher_8_1_10.npy b/experiments/encoders_repo/re_indexer_hasher_8_1_10.npy
diff --git a/experiments/encoders_repo/re_indexer_hasher_9_1_10.npy b/experiments/encoders_repo/re_indexer_hasher_9_1_10.npy