trying to attach caltrans district

Author: amandaha8

gtfs_digest/31_download_prep_data.ipynb

@@ -10,8 +10,8 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
-   "id": "cdb02606-2d1c-4090-9af7-1aee7fed44a7",
+   "execution_count": 1,
+   "id": "a78470b0-9a68-4bc3-8100-d9a2679c9905",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -24,8 +24,26 @@
     "from calitp_data_analysis.gcs_geopandas import GCSGeoPandas\n",
     "\n",
     "import pandas as pd\n",
-    "from functools import cache\n",
-    "\n",
+    "from functools import cache"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "4792e66b-950f-4df4-977e-5694d69e3f83",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from shared_utils import gtfs_utils_v2, portfolio_utils, publish_utils, geo_utils"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "cdb02606-2d1c-4090-9af7-1aee7fed44a7",
+   "metadata": {},
+   "outputs": [],
+   "source": [
     "import publish_public_data\n",
     "from calitp_data_analysis import geography_utils\n",
     "from shared_utils import gtfs_utils_v2, portfolio_utils, publish_utils, geo_utils\n",
@@ -34,7 +52,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 4,
    "id": "5705a199-5cc8-4c64-956c-ba0f62dcdf16",
    "metadata": {
     "tags": []
@@ -49,7 +67,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 5,
    "id": "4b87f16a-943a-49c9-b524-e8cabaa579d6",
    "metadata": {
     "tags": []
@@ -61,7 +79,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 6,
    "id": "691e4dcc-0067-48a3-ae7d-9a956da31074",
    "metadata": {
     "tags": []
@@ -75,7 +93,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 7,
    "id": "730d1e7d-1e07-4adc-8888-8f697641a422",
    "metadata": {
     "tags": []
@@ -326,7 +344,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 8,
    "id": "4c0b9783-97d1-4031-87d2-eecd546156db",
    "metadata": {
     "tags": []
@@ -360,7 +378,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 9,
    "id": "44e48d17-738a-40f5-86cd-943e4d7536c2",
    "metadata": {
     "tags": []
@@ -413,30 +431,92 @@
     "        how=\"left\",\n",
     "    ).drop(columns = [\"month_first_day\"])\n",
     "\n",
-    "    m1 = geo_utils.convert_to_gdf(m1, \"pt_array\", \"line\")\n",
-    "    \"\"\"\n",
+    "    \n",
     "    # Convert the geometry to line\n",
+    "    m1 = geo_utils.convert_to_gdf(m1, \"pt_array\", \"line\")\n",
     "    \n",
     "    m1 = gcs_geopandas().geo_data_frame_to_parquet(\n",
     "        m1,\n",
     "        f\"{GTFS_DATA_DICT.gcs_paths.DIGEST_GCS}processed/fct_monthly_routes_{download_date}.parquet\"\n",
     "    )\n",
-    "    \"\"\"\n",
     "    return m1"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 10,
    "id": "ea7d4e28-78d8-4c69-ad1c-a22cc69eec1c",
    "metadata": {
     "tags": []
    },
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "\n",
+      "        SELECT \n",
+      "            *\n",
+      "        FROM `cal-itp-data-infra-staging`.`tiffany_mart_gtfs_rollup`.`fct_monthly_routes`\n",
+      "        WHERE month_first_day >=  DATE('2025-01-01')\n",
+      "    \n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.11/site-packages/pandas_gbq/gbq.py:38: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81.\n",
+      "  import pkg_resources  # noqa\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Downloading: 100%|\u001b[32m██████████\u001b[0m|\n",
+      "download time: 0:10:16.570742\n"
+     ]
+    }
+   ],
    "source": [
     "fct_operator_hourly_summary = clean_fct_monthly_routes(date_str)"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "acd7c563-e794-4291-8ebb-b387ff6a1238",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(76342, 11)"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "fct_operator_hourly_summary.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "id": "4584dec1-88a0-44bb-8e66-0a4a6f32e4f1",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fct_operator_hourly_summary = gcs_geopandas().geo_data_frame_to_parquet(\n",
+    "        fct_operator_hourly_summary,\n",
+    "        f\"{GTFS_DATA_DICT.gcs_paths.DIGEST_GCS}processed/fct_monthly_routes_{date_str}.parquet\"\n",
+    "    )"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "f282ac48-8e51-4337-889e-ee9a381aba72",