diff --git a/analysis/analysis2.ipynb b/analysis/analysis2.ipynb
index 94bfbca..6e28480 100644
--- a/analysis/analysis2.ipynb
+++ b/analysis/analysis2.ipynb
@@ -298,7 +298,7 @@
"Model: OLS Adj. R-squared: -0.000\n",
"Method: Least Squares F-statistic: 0.2970\n",
"Date: Fri, 09 Aug 2024 Prob (F-statistic): 0.586\n",
- "Time: 15:02:29 Log-Likelihood: -1.0577e+05\n",
+ "Time: 15:56:51 Log-Likelihood: -1.0577e+05\n",
"No. Observations: 16890 AIC: 2.116e+05\n",
"Df Residuals: 16888 BIC: 2.116e+05\n",
"Df Model: 1 \n",
@@ -739,7 +739,7 @@
},
{
"cell_type": "code",
- "execution_count": 20,
+ "execution_count": 19,
"metadata": {},
"outputs": [
{
@@ -759,7 +759,7 @@
"Name: count, dtype: int64"
]
},
- "execution_count": 20,
+ "execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
@@ -774,7 +774,7 @@
},
{
"cell_type": "code",
- "execution_count": 21,
+ "execution_count": 20,
"metadata": {},
"outputs": [
{
@@ -794,7 +794,7 @@
"Name: count, dtype: int64"
]
},
- "execution_count": 21,
+ "execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
@@ -809,7 +809,7 @@
},
{
"cell_type": "code",
- "execution_count": 22,
+ "execution_count": 21,
"metadata": {},
"outputs": [
{
@@ -830,7 +830,7 @@
"Name: count, Length: 275, dtype: int64"
]
},
- "execution_count": 22,
+ "execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
@@ -845,7 +845,7 @@
},
{
"cell_type": "code",
- "execution_count": 23,
+ "execution_count": 22,
"metadata": {},
"outputs": [
{
@@ -888,7 +888,7 @@
"Name: count, dtype: int64"
]
},
- "execution_count": 23,
+ "execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
@@ -903,7 +903,7 @@
},
{
"cell_type": "code",
- "execution_count": 25,
+ "execution_count": 23,
"metadata": {},
"outputs": [
{
@@ -998,7 +998,7 @@
"08001008538 75.181714 "
]
},
- "execution_count": 25,
+ "execution_count": 23,
"metadata": {},
"output_type": "execute_result"
}
@@ -1018,7 +1018,7 @@
},
{
"cell_type": "code",
- "execution_count": 26,
+ "execution_count": 24,
"metadata": {},
"outputs": [
{
@@ -1095,7 +1095,7 @@
"percent_white -0.067320 1.000000 "
]
},
- "execution_count": 26,
+ "execution_count": 24,
"metadata": {},
"output_type": "execute_result"
}
@@ -1110,7 +1110,7 @@
},
{
"cell_type": "code",
- "execution_count": 27,
+ "execution_count": 25,
"metadata": {},
"outputs": [
{
@@ -1175,7 +1175,7 @@
},
{
"cell_type": "code",
- "execution_count": 29,
+ "execution_count": 26,
"metadata": {},
"outputs": [
{
@@ -1188,7 +1188,7 @@
"dtype: int64"
]
},
- "execution_count": 29,
+ "execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
@@ -1203,7 +1203,7 @@
},
{
"cell_type": "code",
- "execution_count": 30,
+ "execution_count": 27,
"metadata": {},
"outputs": [
{
@@ -1216,7 +1216,7 @@
"Model: OLS Adj. R-squared: 0.008\n",
"Method: Least Squares F-statistic: 1.403\n",
"Date: Fri, 09 Aug 2024 Prob (F-statistic): 0.244\n",
- "Time: 15:08:59 Log-Likelihood: -1091.6\n",
+ "Time: 15:56:53 Log-Likelihood: -1091.6\n",
"No. Observations: 157 AIC: 2191.\n",
"Df Residuals: 153 BIC: 2203.\n",
"Df Model: 3 \n",
@@ -1260,7 +1260,7 @@
},
{
"cell_type": "code",
- "execution_count": 31,
+ "execution_count": 28,
"metadata": {},
"outputs": [
{
@@ -1337,7 +1337,7 @@
"percent_white -0.067320 1.000000 "
]
},
- "execution_count": 31,
+ "execution_count": 28,
"metadata": {},
"output_type": "execute_result"
}
@@ -1352,10 +1352,549 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 30,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " Initial Report Date | \n",
+ " Report Year | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " | 0 | \n",
+ " 2014-06-18 | \n",
+ " 2014 | \n",
+ "
\n",
+ " \n",
+ " | 1 | \n",
+ " 2014-06-19 | \n",
+ " 2014 | \n",
+ "
\n",
+ " \n",
+ " | 2 | \n",
+ " 2014-06-19 | \n",
+ " 2014 | \n",
+ "
\n",
+ " \n",
+ " | 3 | \n",
+ " 2014-06-20 | \n",
+ " 2014 | \n",
+ "
\n",
+ " \n",
+ " | 4 | \n",
+ " 2014-06-23 | \n",
+ " 2014 | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " Report Year | \n",
+ " Number of Spills | \n",
+ " Average Spills (Jan-Aug) | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " | 0 | \n",
+ " 2015 | \n",
+ " 1145 | \n",
+ " 143.125 | \n",
+ "
\n",
+ " \n",
+ " | 1 | \n",
+ " 2016 | \n",
+ " 900 | \n",
+ " 112.500 | \n",
+ "
\n",
+ " \n",
+ " | 2 | \n",
+ " 2017 | \n",
+ " 1100 | \n",
+ " 137.500 | \n",
+ "
\n",
+ " \n",
+ " | 3 | \n",
+ " 2018 | \n",
+ " 1069 | \n",
+ " 133.625 | \n",
+ "
\n",
+ " \n",
+ " | 4 | \n",
+ " 2019 | \n",
+ " 1056 | \n",
+ " 132.000 | \n",
+ "
\n",
+ " \n",
+ " | 5 | \n",
+ " 2020 | \n",
+ " 827 | \n",
+ " 103.375 | \n",
+ "
\n",
+ " \n",
+ " | 6 | \n",
+ " 2021 | \n",
+ " 1143 | \n",
+ " 142.875 | \n",
+ "
\n",
+ " \n",
+ " | 7 | \n",
+ " 2022 | \n",
+ " 1465 | \n",
+ " 183.125 | \n",
+ "
\n",
+ " \n",
+ " | 8 | \n",
+ " 2023 | \n",
+ " 1420 | \n",
+ " 177.500 | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "import matplotlib.pyplot as plt\n",
+ "import seaborn as sns\n",
+ "\n",
+ "# Plot the average number of spills per year (January-August) for 2015-2023\n",
+ "plt.figure(figsize=(12, 6))\n",
+ "sns.lineplot(x='Report Year', y='Average Spills (Jan-Aug)', data=spills_by_year_filtered, marker='o')\n",
+ "plt.title('Average Number of Spills Reported by Year (January-August, 2015-2023)')\n",
+ "plt.xlabel('Year')\n",
+ "plt.ylabel('Average Number of Spills')\n",
+ "plt.grid(True)\n",
+ "plt.show()\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Reporting Delay by Year"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 33,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " Report Year | \n",
+ " Average Report Delay (Days) | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " | 0 | \n",
+ " 2014 | \n",
+ " 5.991485 | \n",
+ "
\n",
+ " \n",
+ " | 1 | \n",
+ " 2015 | \n",
+ " 31.561963 | \n",
+ "
\n",
+ " \n",
+ " | 2 | \n",
+ " 2016 | \n",
+ " 8.813953 | \n",
+ "
\n",
+ " \n",
+ " | 3 | \n",
+ " 2017 | \n",
+ " 4.448169 | \n",
+ "
\n",
+ " \n",
+ " | 4 | \n",
+ " 2018 | \n",
+ " 2.939987 | \n",
+ "
\n",
+ " \n",
+ " | 5 | \n",
+ " 2019 | \n",
+ " 16.108612 | \n",
+ "
\n",
+ " \n",
+ " | 6 | \n",
+ " 2020 | \n",
+ " 12.212947 | \n",
+ "
\n",
+ " \n",
+ " | 7 | \n",
+ " 2021 | \n",
+ " 3.074157 | \n",
+ "
\n",
+ " \n",
+ " | 8 | \n",
+ " 2022 | \n",
+ " 7.804287 | \n",
+ "
\n",
+ " \n",
+ " | 9 | \n",
+ " 2023 | \n",
+ " 6.325469 | \n",
+ "
\n",
+ " \n",
+ " | 10 | \n",
+ " 2024 | \n",
+ " 12.123333 | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "# Plot the average reporting delay by year\n",
+ "plt.figure(figsize=(12, 6))\n",
+ "sns.lineplot(x='Report Year', y='Average Report Delay (Days)', data=report_delay_by_year, marker='o')\n",
+ "plt.title('Average Reporting Delay by Year')\n",
+ "plt.xlabel('Year')\n",
+ "plt.ylabel('Average Report Delay (Days)')\n",
+ "plt.grid(True)\n",
+ "plt.show()\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 41,
"metadata": {},
"outputs": [],
- "source": []
+ "source": [
+ "# Filter the data for years 2015 to 2023\n",
+ "spill_gdf_filtered = spills_gdf[(spills_gdf['Report Year'] >= 2015) & (spills_gdf['Report Year'] <= 2023)]\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 42,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " Report Year | \n",
+ " Average Spills per Month | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " | 0 | \n",
+ " 2015 | \n",
+ " 143.125 | \n",
+ "
\n",
+ " \n",
+ " | 1 | \n",
+ " 2016 | \n",
+ " 112.500 | \n",
+ "
\n",
+ " \n",
+ " | 2 | \n",
+ " 2017 | \n",
+ " 137.500 | \n",
+ "
\n",
+ " \n",
+ " | 3 | \n",
+ " 2018 | \n",
+ " 133.625 | \n",
+ "
\n",
+ " \n",
+ " | 4 | \n",
+ " 2019 | \n",
+ " 132.000 | \n",
+ "
\n",
+ " \n",
+ " | 5 | \n",
+ " 2020 | \n",
+ " 103.375 | \n",
+ "
\n",
+ " \n",
+ " | 6 | \n",
+ " 2021 | \n",
+ " 142.875 | \n",
+ "
\n",
+ " \n",
+ " | 7 | \n",
+ " 2022 | \n",
+ " 183.125 | \n",
+ "
\n",
+ " \n",
+ " | 8 | \n",
+ " 2023 | \n",
+ " 177.500 | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "