save changes

yul19079 · Feb 1, 2024 · 3f4624f · 3f4624f
1 parent 28c5478
commit 3f4624f
Show file tree

Hide file tree

Showing 2 changed files with 705 additions and 15 deletions.
diff --git a/EDA.ipynb b/EDA.ipynb
@@ -29,32 +29,36 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [],
    "source": [
-    "raw_data = pd.read_pickle(\"./data/raw_feature.pkl\")"
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "raw_data = pd.read_pickle(\"./data/raw_feature.pkl\")\n",
+    "X_train = pd.read_pickle(\"./data/X_train.pkl\")\n",
+    "y_train = pd.read_pickle(\"./data/y_train.pkl\")\n",
+    "X_test = pd.read_pickle(\"./data/X_test.pkl\")\n",
+    "y_test = pd.read_pickle(\"./data/y_test.pkl\")\n",
+    "\n",
+    "\n",
+    "trip_avg_in = y_train.mean(axis = 0)[0]\n",
+    "trip_avg_out = y_train.mean(axis=0)[1]"
    ]
   },
   {
    "cell_type": "code",
    "execution_count": 4,
    "metadata": {},
    "outputs": [],
-   "source": [
-    "X_train = pd.read_pickle(\"./data/X_train.pkl\")\n",
-    "y_train = pd.read_pickle(\"./data/y_train.pkl\")"
-   ]
+   "source": []
   },
   {
    "cell_type": "code",
    "execution_count": 5,
    "metadata": {},
    "outputs": [],
-   "source": [
-    "trip_avg_in = y_train.mean(axis = 0)[0]\n",
-    "trip_avg_out = y_train.mean(axis=0)[1]"
-   ]
+   "source": []
   },
   {
    "cell_type": "code",
@@ -64,7 +68,7 @@
     {
      "data": {
       "text/plain": [
-       "<matplotlib.image.AxesImage at 0x1ed9fc2bd50>"
+       "<matplotlib.image.AxesImage at 0x17de86c4190>"
       ]
      },
      "execution_count": 6,
@@ -95,7 +99,7 @@
     {
      "data": {
       "text/plain": [
-       "<matplotlib.image.AxesImage at 0x1ed9fcdbd50>"
+       "<matplotlib.image.AxesImage at 0x17de87647d0>"
       ]
      },
      "execution_count": 7,
@@ -417,7 +421,7 @@
    "outputs": [],
    "source": [
     "import sys\n",
-    "file_path = \"pearson_output.txt\"\n",
+    "file_path = \"pearson_output_filter0.txt\"\n",
     "\n",
     "with open(file_path,\"w\") as file:\n",
     "    original_stdout = sys.stdout\n",
@@ -427,7 +431,7 @@
     "    feature_name_list\n",
     "    for feature_name in feature_name_list:\n",
     "        temp_feature = get_feature_in_2d(raw_data,feature_name)\n",
-    "        cal_pearson(temp_feature,trip_avg_in,log_text=feature_name)\n",
+    "        cal_pearson(temp_feature,trip_avg_in,log_text=feature_name,filter_zero=True)\n",
     "        print(\"\")\n",
     "    sys.stdout = original_stdout"
    ]
@@ -451,7 +455,7 @@
     {
      "data": {
       "text/plain": [
-       "<matplotlib.image.AxesImage at 0x1edd1903d50>"
+       "<matplotlib.image.AxesImage at 0x17dff3d47d0>"
       ]
      },
      "execution_count": 12,