diff --git a/notebooks/format_results.ipynb b/notebooks/format_results.ipynb
new file mode 100644
index 0000000..5e87ccd
--- /dev/null
+++ b/notebooks/format_results.ipynb
@@ -0,0 +1,833 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "id": "ecedb1c0-d2cc-44d5-aa6e-9de150509d6c",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "import glob\n",
+ "import pandas as pd\n",
+ "import seaborn as sns\n",
+ "import matplotlib.pyplot as plt\n",
+ "import matplotlib.image as mpimg\n",
+ "\n",
+ "sns.set_theme()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 66,
+ "id": "b1ba7767-0e5d-482e-a138-e3a2542c3a2f",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "param = 'tn'"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 67,
+ "id": "c13f310a-9569-4b83-b26b-9c26ceb4a39a",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "versions = [f'{param}_model_v{i}' for i in range(1, 5)]"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 68,
+ "id": "9fc471e2-0162-4260-92e6-f7f2b2b69865",
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "['tn_model_v1', 'tn_model_v2', 'tn_model_v3', 'tn_model_v4']"
+ ]
+ },
+ "execution_count": 68,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "versions"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 69,
+ "id": "0c95583c-1243-4ba2-84ad-8939d9a96c73",
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "{'tn_model_v1': 0.9,\n",
+ " 'tn_model_v2': 0.8,\n",
+ " 'tn_model_v3': 0.7,\n",
+ " 'tn_model_v4': 0.6}"
+ ]
+ },
+ "execution_count": 69,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "threshold_dict = dict(zip(versions, [0.9, 0.8, 0.7, 0.6]))\n",
+ "threshold_dict"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 70,
+ "id": "a1757af3-148c-41a8-a7a7-4b88ea2516f0",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "files = [f'D:/est_water_qual/model/{version}/{version}_results.csv' for version in versions]"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 71,
+ "id": "6aa5b0f7-43ca-4667-ad18-ab11276348b1",
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "['D:/est_water_qual/model/tn_model_v1/tn_model_v1_results.csv',\n",
+ " 'D:/est_water_qual/model/tn_model_v2/tn_model_v2_results.csv',\n",
+ " 'D:/est_water_qual/model/tn_model_v3/tn_model_v3_results.csv',\n",
+ " 'D:/est_water_qual/model/tn_model_v4/tn_model_v4_results.csv']"
+ ]
+ },
+ "execution_count": 71,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "files"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 72,
+ "id": "cba86bd0-0930-41ee-83e1-837bfa0bcfc4",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "results = pd.DataFrame({'Attribute': pd.read_csv(files[0])['Attribute']})"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 73,
+ "id": "83c75ff6-36ee-46a3-be94-6fc4b8e494c7",
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "
\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " Attribute | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " 0 | \n",
+ " n_features | \n",
+ "
\n",
+ " \n",
+ " 1 | \n",
+ " test_size | \n",
+ "
\n",
+ " \n",
+ " 2 | \n",
+ " n_samples_train | \n",
+ "
\n",
+ " \n",
+ " 3 | \n",
+ " n_samples_test | \n",
+ "
\n",
+ " \n",
+ " 4 | \n",
+ " r2_train | \n",
+ "
\n",
+ " \n",
+ " 5 | \n",
+ " r2_test | \n",
+ "
\n",
+ " \n",
+ " 6 | \n",
+ " oob_score | \n",
+ "
\n",
+ " \n",
+ " 7 | \n",
+ " nse_test | \n",
+ "
\n",
+ " \n",
+ " 8 | \n",
+ " pbias_test | \n",
+ "
\n",
+ " \n",
+ " 9 | \n",
+ " max_depth | \n",
+ "
\n",
+ " \n",
+ " 10 | \n",
+ " max_features | \n",
+ "
\n",
+ " \n",
+ " 11 | \n",
+ " min_samples_leaf | \n",
+ "
\n",
+ " \n",
+ " 12 | \n",
+ " min_samples_split | \n",
+ "
\n",
+ " \n",
+ " 13 | \n",
+ " n_estimators | \n",
+ "
\n",
+ " \n",
+ " 14 | \n",
+ " bootstrap | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
"
+ ],
+ "text/plain": [
+ " Attribute\n",
+ "0 n_features\n",
+ "1 test_size\n",
+ "2 n_samples_train\n",
+ "3 n_samples_test\n",
+ "4 r2_train\n",
+ "5 r2_test\n",
+ "6 oob_score\n",
+ "7 nse_test\n",
+ "8 pbias_test\n",
+ "9 max_depth\n",
+ "10 max_features\n",
+ "11 min_samples_leaf\n",
+ "12 min_samples_split\n",
+ "13 n_estimators\n",
+ "14 bootstrap"
+ ]
+ },
+ "execution_count": 73,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "results"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 74,
+ "id": "726914fd-5a37-4c69-99ec-c9b69c8cddf7",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "for file, version in zip(files, versions):\n",
+ " df = pd.read_csv(file)\n",
+ " results = results.merge(df, how='outer', on='Attribute')"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 75,
+ "id": "a45dbcc7-4914-4a84-b803-72b87219bbd7",
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " Attribute | \n",
+ " TN_MODEL_V1 | \n",
+ " TN_MODEL_V2 | \n",
+ " TN_MODEL_V3 | \n",
+ " TN_MODEL_V4 | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " 0 | \n",
+ " n_features | \n",
+ " 62 | \n",
+ " 55 | \n",
+ " 47 | \n",
+ " 38 | \n",
+ "
\n",
+ " \n",
+ " 1 | \n",
+ " test_size | \n",
+ " 0.3 | \n",
+ " 0.3 | \n",
+ " 0.3 | \n",
+ " 0.3 | \n",
+ "
\n",
+ " \n",
+ " 2 | \n",
+ " n_samples_train | \n",
+ " 328 | \n",
+ " 328 | \n",
+ " 328 | \n",
+ " 328 | \n",
+ "
\n",
+ " \n",
+ " 3 | \n",
+ " n_samples_test | \n",
+ " 141 | \n",
+ " 141 | \n",
+ " 141 | \n",
+ " 141 | \n",
+ "
\n",
+ " \n",
+ " 4 | \n",
+ " r2_train | \n",
+ " 0.8768110787259299 | \n",
+ " 0.9370844249679031 | \n",
+ " 0.9174176077581037 | \n",
+ " 0.9277804571934218 | \n",
+ "
\n",
+ " \n",
+ " 5 | \n",
+ " r2_test | \n",
+ " 0.7914042073068632 | \n",
+ " 0.8345873204768984 | \n",
+ " 0.8205758597605688 | \n",
+ " 0.8325944480437459 | \n",
+ "
\n",
+ " \n",
+ " 6 | \n",
+ " oob_score | \n",
+ " 0.7011273201967085 | \n",
+ " 0.7268873570430123 | \n",
+ " 0.7136227756184379 | \n",
+ " 0.7378667449603072 | \n",
+ "
\n",
+ " \n",
+ " 7 | \n",
+ " nse_test | \n",
+ " 0.7914042073068632 | \n",
+ " 0.8345873204768984 | \n",
+ " 0.8205758597605688 | \n",
+ " 0.8325944480437459 | \n",
+ "
\n",
+ " \n",
+ " 8 | \n",
+ " pbias_test | \n",
+ " -0.005326843287056666 | \n",
+ " 0.034952618333772785 | \n",
+ " 0.018924951828989685 | \n",
+ " 0.01821177573840633 | \n",
+ "
\n",
+ " \n",
+ " 9 | \n",
+ " max_depth | \n",
+ " 90 | \n",
+ " 30 | \n",
+ " 30 | \n",
+ " 60 | \n",
+ "
\n",
+ " \n",
+ " 10 | \n",
+ " max_features | \n",
+ " sqrt | \n",
+ " sqrt | \n",
+ " sqrt | \n",
+ " sqrt | \n",
+ "
\n",
+ " \n",
+ " 11 | \n",
+ " min_samples_leaf | \n",
+ " 4 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ "
\n",
+ " \n",
+ " 12 | \n",
+ " min_samples_split | \n",
+ " 5 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 5 | \n",
+ "
\n",
+ " \n",
+ " 13 | \n",
+ " n_estimators | \n",
+ " 30 | \n",
+ " 30 | \n",
+ " 30 | \n",
+ " 60 | \n",
+ "
\n",
+ " \n",
+ " 14 | \n",
+ " bootstrap | \n",
+ " True | \n",
+ " True | \n",
+ " True | \n",
+ " True | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
"
+ ],
+ "text/plain": [
+ " Attribute TN_MODEL_V1 TN_MODEL_V2 \\\n",
+ "0 n_features 62 55 \n",
+ "1 test_size 0.3 0.3 \n",
+ "2 n_samples_train 328 328 \n",
+ "3 n_samples_test 141 141 \n",
+ "4 r2_train 0.8768110787259299 0.9370844249679031 \n",
+ "5 r2_test 0.7914042073068632 0.8345873204768984 \n",
+ "6 oob_score 0.7011273201967085 0.7268873570430123 \n",
+ "7 nse_test 0.7914042073068632 0.8345873204768984 \n",
+ "8 pbias_test -0.005326843287056666 0.034952618333772785 \n",
+ "9 max_depth 90 30 \n",
+ "10 max_features sqrt sqrt \n",
+ "11 min_samples_leaf 4 2 \n",
+ "12 min_samples_split 5 2 \n",
+ "13 n_estimators 30 30 \n",
+ "14 bootstrap True True \n",
+ "\n",
+ " TN_MODEL_V3 TN_MODEL_V4 \n",
+ "0 47 38 \n",
+ "1 0.3 0.3 \n",
+ "2 328 328 \n",
+ "3 141 141 \n",
+ "4 0.9174176077581037 0.9277804571934218 \n",
+ "5 0.8205758597605688 0.8325944480437459 \n",
+ "6 0.7136227756184379 0.7378667449603072 \n",
+ "7 0.8205758597605688 0.8325944480437459 \n",
+ "8 0.018924951828989685 0.01821177573840633 \n",
+ "9 30 60 \n",
+ "10 sqrt sqrt \n",
+ "11 2 2 \n",
+ "12 2 5 \n",
+ "13 30 60 \n",
+ "14 True True "
+ ]
+ },
+ "execution_count": 75,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "results"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 76,
+ "id": "299b5a6a-bec3-4a29-8af4-9ddf14ce8aa8",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "for version in versions:\n",
+ " for i in range(4, 9):\n",
+ " results[version.upper()].iloc[i] = round(float(results[version.upper()].iloc[i]), 3)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 77,
+ "id": "9aa781e8-0c28-429f-9dbd-aafaefd04c61",
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " Attribute | \n",
+ " TN_MODEL_V1 | \n",
+ " TN_MODEL_V2 | \n",
+ " TN_MODEL_V3 | \n",
+ " TN_MODEL_V4 | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " 0 | \n",
+ " n_features | \n",
+ " 62 | \n",
+ " 55 | \n",
+ " 47 | \n",
+ " 38 | \n",
+ "
\n",
+ " \n",
+ " 1 | \n",
+ " test_size | \n",
+ " 0.3 | \n",
+ " 0.3 | \n",
+ " 0.3 | \n",
+ " 0.3 | \n",
+ "
\n",
+ " \n",
+ " 2 | \n",
+ " n_samples_train | \n",
+ " 328 | \n",
+ " 328 | \n",
+ " 328 | \n",
+ " 328 | \n",
+ "
\n",
+ " \n",
+ " 3 | \n",
+ " n_samples_test | \n",
+ " 141 | \n",
+ " 141 | \n",
+ " 141 | \n",
+ " 141 | \n",
+ "
\n",
+ " \n",
+ " 4 | \n",
+ " r2_train | \n",
+ " 0.877 | \n",
+ " 0.937 | \n",
+ " 0.917 | \n",
+ " 0.928 | \n",
+ "
\n",
+ " \n",
+ " 5 | \n",
+ " r2_test | \n",
+ " 0.791 | \n",
+ " 0.835 | \n",
+ " 0.821 | \n",
+ " 0.833 | \n",
+ "
\n",
+ " \n",
+ " 6 | \n",
+ " oob_score | \n",
+ " 0.701 | \n",
+ " 0.727 | \n",
+ " 0.714 | \n",
+ " 0.738 | \n",
+ "
\n",
+ " \n",
+ " 7 | \n",
+ " nse_test | \n",
+ " 0.791 | \n",
+ " 0.835 | \n",
+ " 0.821 | \n",
+ " 0.833 | \n",
+ "
\n",
+ " \n",
+ " 8 | \n",
+ " pbias_test | \n",
+ " -0.005 | \n",
+ " 0.035 | \n",
+ " 0.019 | \n",
+ " 0.018 | \n",
+ "
\n",
+ " \n",
+ " 9 | \n",
+ " max_depth | \n",
+ " 90 | \n",
+ " 30 | \n",
+ " 30 | \n",
+ " 60 | \n",
+ "
\n",
+ " \n",
+ " 10 | \n",
+ " max_features | \n",
+ " sqrt | \n",
+ " sqrt | \n",
+ " sqrt | \n",
+ " sqrt | \n",
+ "
\n",
+ " \n",
+ " 11 | \n",
+ " min_samples_leaf | \n",
+ " 4 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ "
\n",
+ " \n",
+ " 12 | \n",
+ " min_samples_split | \n",
+ " 5 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 5 | \n",
+ "
\n",
+ " \n",
+ " 13 | \n",
+ " n_estimators | \n",
+ " 30 | \n",
+ " 30 | \n",
+ " 30 | \n",
+ " 60 | \n",
+ "
\n",
+ " \n",
+ " 14 | \n",
+ " bootstrap | \n",
+ " True | \n",
+ " True | \n",
+ " True | \n",
+ " True | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
"
+ ],
+ "text/plain": [
+ " Attribute TN_MODEL_V1 TN_MODEL_V2 TN_MODEL_V3 TN_MODEL_V4\n",
+ "0 n_features 62 55 47 38\n",
+ "1 test_size 0.3 0.3 0.3 0.3\n",
+ "2 n_samples_train 328 328 328 328\n",
+ "3 n_samples_test 141 141 141 141\n",
+ "4 r2_train 0.877 0.937 0.917 0.928\n",
+ "5 r2_test 0.791 0.835 0.821 0.833\n",
+ "6 oob_score 0.701 0.727 0.714 0.738\n",
+ "7 nse_test 0.791 0.835 0.821 0.833\n",
+ "8 pbias_test -0.005 0.035 0.019 0.018\n",
+ "9 max_depth 90 30 30 60\n",
+ "10 max_features sqrt sqrt sqrt sqrt\n",
+ "11 min_samples_leaf 4 2 2 2\n",
+ "12 min_samples_split 5 2 2 5\n",
+ "13 n_estimators 30 30 30 60\n",
+ "14 bootstrap True True True True"
+ ]
+ },
+ "execution_count": 77,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "results"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 78,
+ "id": "31255d10-61b4-4c4b-adef-b6e4176fd58e",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "results = results.drop(results.index[7:9]).reset_index(drop=True)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 79,
+ "id": "a8d89d9c-e265-47e7-bd81-0a5cf7d429fc",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "caption = f'''Results of the four model versions used for {param.upper()} prediction along with hyperparameters derived from the RandomizedSearchCV \n",
+ "algorithm.'''\n",
+ "results.to_latex(\n",
+ " f'D:/est_water_qual/model/{param}_results.tex', index=False, longtable=True, caption=caption, label=f'table:{param}_results'\n",
+ ")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 85,
+ "id": "5d5e4461-956b-4a9a-833f-3fb91a3aef14",
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "