abandono_trat_tuberculose_analise.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d5a8c034",
   "metadata": {},
   "source": [
    "# CLASSIFICADOR DE ABANDONO DO TRATAMENTO DA TURBECULOSE\n",
    "O dataset utilizado nesse projeto contém casos de tuberculose pulmonar reportados ao Sistema de Informação de Agravos de Notificação (SINAN) entre 2015 e 2022. Dessa forma, faremos uma análise dos dados para tentar entender certos padrões e quais fatores estão mais associados ao caso da não-finalização do tratamento da tuberculose, ou seja, do abandono do tratamento. Após a analise, treinaremos um modelo classificador."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f523b404-10fa-4abd-bf6f-c7aeea8bf2fb",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.experimental import enable_iterative_imputer\n",
    "from sklearn.impute import IterativeImputer\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "import plotly.express as px"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae75df4c-04a9-4245-9f55-1376c070a902",
   "metadata": {},
   "source": [
    "## Carregando o dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cd68b233-8852-4915-bb7d-8164dffea496",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_32693/2336186354.py:1: DtypeWarning: Columns (22,24,29,30,32,49,54,60,76,93) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  df = pd.read_csv('casos_tuberculose_2015-2022.csv')\n",
      "/tmp/ipykernel_32693/2336186354.py:5: FutureWarning: Downcasting behavior in `replace` is deprecated and will be removed in a future version. To retain the old behavior, explicitly call `result.infer_objects(copy=False)`. To opt-in to the future behavior, set `pd.set_option('future.no_silent_downcasting', True)`\n",
      "  df['reg_not'] = df['reg_not'].replace({'North': 0, 'NorthEast': 1, 'South': 2, 'SouthEast': 3, 'Central-west': 4})\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>reg_not</th>\n",
       "      <th>cs_sexo</th>\n",
       "      <th>cs_raca</th>\n",
       "      <th>cs_escol_n</th>\n",
       "      <th>tratamento</th>\n",
       "      <th>raiox_tora</th>\n",
       "      <th>agravaids</th>\n",
       "      <th>agravalcoo</th>\n",
       "      <th>agravdiabe</th>\n",
       "      <th>agravdoenc</th>\n",
       "      <th>hiv</th>\n",
       "      <th>tratsup_at</th>\n",
       "      <th>benef_gov</th>\n",
       "      <th>agravdroga</th>\n",
       "      <th>agravtabac</th>\n",
       "      <th>idade</th>\n",
       "      <th>abandono</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>37.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   reg_not  cs_sexo  cs_raca  cs_escol_n  tratamento  raiox_tora  agravaids  \\\n",
       "0      4.0      1.0      1.0         1.0         1.0         0.0        0.0   \n",
       "1      1.0      1.0      1.0         0.0         1.0         1.0        NaN   \n",
       "2      2.0      1.0      1.0         0.0         1.0         2.0        NaN   \n",
       "3      2.0      1.0      1.0         1.0         1.0         1.0        0.0   \n",
       "4      3.0      1.0      1.0         1.0         0.0         1.0        NaN   \n",
       "\n",
       "   agravalcoo  agravdiabe  agravdoenc  hiv  tratsup_at  benef_gov  agravdroga  \\\n",
       "0         1.0         0.0         0.0  0.0         1.0        NaN         0.0   \n",
       "1         1.0         NaN         NaN  2.0         0.0        0.0         NaN   \n",
       "2         NaN         NaN         NaN  2.0         0.0        0.0         1.0   \n",
       "3         1.0         0.0         0.0  0.0         NaN        0.0         1.0   \n",
       "4         1.0         NaN         NaN  0.0         1.0        1.0         1.0   \n",
       "\n",
       "   agravtabac  idade  abandono  \n",
       "0         1.0   60.0         0  \n",
       "1         NaN   37.0         1  \n",
       "2         NaN   22.0         1  \n",
       "3         1.0   37.0         0  \n",
       "4         1.0   22.0         1  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# df = pd.read_csv('casos_tuberculose_2015-2022.csv')\n",
    "# df = df.drop(['tp_not','id_agravo','dt_notific','nu_ano','id_municip','id_regiona','dt_diag','ano_nasc','nu_idade_n','sg_uf','id_mn_resi','id_rg_resi','id_pais','nduplic_n','in_vincula','dt_digita','dt_transus','dt_transdm','dt_transsm','dt_transrm','dt_transrs','dt_transse','cs_flxret','flxrecebi','migrado_w','id_ocupa_n','institucio','teste_tube','extrapu1_n','extrapu2_n','extrapul_o','agravoutra','agravoutde','bacilosc_e','bacilos_e2','bacilosc_o','cultura_ou','histopatol','dt_inic_tr','rifampicin','isoniazida','etambutol','estreptomi','pirazinami','etionamida','outras','outras_des','nu_contato','doenca_tra','sg_uf_at','id_munic_a','dt_noti_at','sg_uf_2','id_munic_2','bacilosc_1','bacilosc_2','bacilosc_3','bacilosc_4','bacilosc_5','bacilosc_6','dt_mudanca','nu_comu_ex','situa_9_m','situa_12_m','dt_encerra','tpuninot','test_molec','test_sensi','ant_retro','bac_apos_6','transf','uf_transf','mun_transf','time_in_SINAN','cultura_es','trat_super','cs_gestant','situa_ence','pop_liber','pop_rua','pop_saude','pop_imig','forma'], axis=1)\n",
    "\n",
    "# df.rename(columns={'sg_uf_not': 'reg_not'}, inplace=True)\n",
    "# df['reg_not'] = df['reg_not'].replace({'North': 0, 'NorthEast': 1, 'South': 2, 'SouthEast': 3, 'Central-west': 4})\n",
    "# df['raiox_tora'] = df['raiox_tora'].replace({np.NaN:2}) # Valores nulos considerados como raio-x não realizado\n",
    "# df['agravaids'] = df['agravaids'].replace({2:0, 3:np.NaN, 9:np.NaN})\n",
    "# df['hiv'] = df['hiv'].replace({np.NaN:2}) # Valores nulos considerados como teste HIV não realizado\n",
    "# df.rename(columns={'outcome': 'abandono'}, inplace=True)\n",
    "# df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc0ccc44-874f-4062-b8d2-9fd8e562e57f",
   "metadata": {},
   "source": [
    "### Lidando com dados faltantes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "92cf6960-d3ad-429c-9bb6-69a0811ddcdf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plt.title('Dados faltantes',fontweight='bold')\n",
    "# ax = sns.heatmap(df.isna().sum().to_frame(),annot=True,fmt='d',cmap='vlag')\n",
    "# ax.set_xlabel('Quantidade faltante')\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11b9afad-3840-4ac9-ac38-4495b04c04cb",
   "metadata": {},
   "source": [
    "Nossa base de dados possui uma grande quantidade de dados faltantes, portanto vamos utilizar o método da imputação iterativa com um classificador de árvore de decisão para preencher os dados, sabendo que os dados nulos são todos categóricos."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2727ff02-73e8-4a64-a492-80acea402228",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Imputador iterativo com um classificador de árvore de decisão\n",
    "# iterative_imputer = IterativeImputer(estimator=DecisionTreeClassifier(), random_state=42)\n",
    "# df_imputed = pd.DataFrame(iterative_imputer.fit_transform(df), columns=df.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8868089",
   "metadata": {},
   "source": [
    "Vamos salvar o dataframe imputado para não precisar rodar o imputador novamente. Usaremos ele para analisar os dados e para treinar o modelo."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "35139262",
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_imputed.to_csv(\"data/tuberculose_imputed.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f49507f7-8376-459d-b35b-99e04aba660b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 243726 entries, 0 to 243725\n",
      "Data columns (total 17 columns):\n",
      " #   Column      Non-Null Count   Dtype\n",
      "---  ------      --------------   -----\n",
      " 0   reg_not     243726 non-null  int8 \n",
      " 1   cs_sexo     243726 non-null  int8 \n",
      " 2   cs_raca     243726 non-null  int8 \n",
      " 3   cs_escol_n  243726 non-null  int8 \n",
      " 4   tratamento  243726 non-null  int8 \n",
      " 5   raiox_tora  243726 non-null  int8 \n",
      " 6   agravaids   243726 non-null  int8 \n",
      " 7   agravalcoo  243726 non-null  int8 \n",
      " 8   agravdiabe  243726 non-null  int8 \n",
      " 9   agravdoenc  243726 non-null  int8 \n",
      " 10  hiv         243726 non-null  int8 \n",
      " 11  tratsup_at  243726 non-null  int8 \n",
      " 12  benef_gov   243726 non-null  int8 \n",
      " 13  agravdroga  243726 non-null  int8 \n",
      " 14  agravtabac  243726 non-null  int8 \n",
      " 15  idade       243726 non-null  int8 \n",
      " 16  abandono    243726 non-null  int8 \n",
      "dtypes: int8(17)\n",
      "memory usage: 4.0 MB\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('data/tuberculose_imputed.csv')\n",
    "# Otimização de memória\n",
    "for column in df.columns:\n",
    "    df[column] = df[column].astype(np.int8)\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "890facde-7941-43f1-94a7-9c27e93dab14",
   "metadata": {},
   "source": [
    "## Exploração dos dados"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc0d3b43",
   "metadata": {},
   "source": [
    "Inicialmente, vamos analisar a proporção de casos de abandono e não-abandono do dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0359633c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "counts = df['abandono'].value_counts()\n",
    "\n",
    "# Plotar o gráfico de pizza\n",
    "plt.figure(figsize=(4, 4))\n",
    "plt.pie(counts, labels=['0', '1'], autopct='%1.1f%%', startangle=90, colors=['lightcoral', 'lightblue'])\n",
    "plt.title('Abandono', fontweight='bold')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4b7e5ef",
   "metadata": {},
   "source": [
    "Percebemos que o dataset é muito desbalanceado, sendo 17% dos casos com desfecho de abandono. Geralmente, modelos de Machine Learning não desempenham bem em casos assim, portanto trataremos disso quando formos treinar os modelos. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5cd2b58",
   "metadata": {},
   "source": [
    "### Análise bivariada"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32ccabf3",
   "metadata": {},
   "source": [
    "Vamos analisar a relação do abandono com cada variável."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb0a24f3",
   "metadata": {},
   "source": [
    "- #### Idade & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f8efbde2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x1500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conts = ['idade']\n",
    "background_color = \"#fafafa\"\n",
    "\n",
    "fig = plt.figure(figsize=(10, 10), dpi=150,facecolor=background_color)\n",
    "gs = fig.add_gridspec(4, 3)\n",
    "gs.update(wspace=0.1, hspace=0.4)\n",
    "\n",
    "\n",
    "plot = 0\n",
    "locals()[\"ax\"+str(plot)] = fig.add_subplot(gs[0, 0])\n",
    "locals()[\"ax\"+str(plot)].set_facecolor(background_color)\n",
    "locals()[\"ax\"+str(plot)].tick_params(axis='y', left=False)\n",
    "locals()[\"ax\"+str(plot)].get_yaxis().set_visible(False)\n",
    "for s in [\"top\",\"right\",\"left\"]:\n",
    "    locals()[\"ax\"+str(plot)].spines[s].set_visible(False)\n",
    "\n",
    "a = df[df['abandono'] == 1]\n",
    "na = df[df['abandono'] == 0]\n",
    "\n",
    "sns.kdeplot(a['idade'], ax=locals()[\"ax\"+str(plot)], color='#0f4c81', fill=True, linewidth=1.5, ec='black',alpha=0.9, zorder=3, legend=False)\n",
    "sns.kdeplot(na['idade'],ax=locals()[\"ax\"+str(plot)], color='#9bb7d4', fill=True, linewidth=1.5, ec='black',alpha=0.9, zorder=3, legend=False)\n",
    "locals()[\"ax\"+str(plot)].grid(which='major', axis='x', zorder=0, color='gray', linestyle=':', dashes=(1,5))\n",
    "plt.legend(['Abandono', 'Não abandono'], loc='upper right', fontsize=8)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "8aa315c2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Média de idade do grupo Abandono: 37.1 anos\n",
      "Média de idade do grupo Cura: 43.1 anos\n"
     ]
    }
   ],
   "source": [
    "mediaAbandono = a['idade'].mean()\n",
    "mediaCura = na['idade'].mean()\n",
    "print(f'Média de idade do grupo Abandono: {mediaAbandono:.1f} anos')\n",
    "print(f'Média de idade do grupo Cura: {mediaCura:.1f} anos')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc0ccd99",
   "metadata": {},
   "source": [
    "Comparando a idade dos grupos de pessoas que abandonaram o tratamento com as que não abandonaram, percecbe-se uma tendência maior das pessoas do primeiro grupo a serem mais jovens."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c4bde1f",
   "metadata": {},
   "source": [
    "- #### Sexo & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a4c09d9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"cs_sexo\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Feminino\", \"Masculino\"]\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d540d7af-c2ce-4aad-a121-39bd709f538b",
   "metadata": {},
   "source": [
    "Nota-se que quase o dobro dos casos de tuberculose ocorre em homens e, percentualmente, há uma chance menor de abandono por mulheres.  \n",
    "13% dos pacientes do sexo feminino abandonaram o tratamento, enquanto 19% do sexo masculino abandonaram. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e11c8ca",
   "metadata": {},
   "source": [
    "- #### Região de notificação & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "99f8c03b",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"reg_not\", color=\"abandono\",width=400, height=400)\n",
    "# Definir os valores e os rótulos desejados\n",
    "tickvals = [0, 1, 2, 3, 4]\n",
    "ticktext = [\"Norte\", \"Nordeste\", \"Sul\", \"Sudeste\", \"Centro-Oeste\"]\n",
    "\n",
    "# Atualiza os eixos x\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4f0051f",
   "metadata": {},
   "source": [
    "Na região Norte, 17% dos casos notificados resultaram em abandono;\n",
    "No Nordeste, também 17%;\n",
    "No Sul, 19%;\n",
    "No Sudeste, 16%;\n",
    "No Centro-Oeste, 19%.  \n",
    "Conclui-se que não há grande diferença na proporção de abandono entre as diferentes regiões.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "170a7aa9",
   "metadata": {},
   "source": [
    "- #### Raça & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "90c8013d",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"cs_raca\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Branco\", \"Preto, Pardo, Amarelo, Indígena\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f047e13",
   "metadata": {},
   "source": [
    "13% das pessoas brancas com tuberculose abandonaram. Das pessoas de raça preta, parda, amarela ou indígena, 18,6% abandonaram. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "388b7b88",
   "metadata": {},
   "source": [
    "- #### Nível de escolaridade & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ebeb3099",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"cs_escol_n\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1, 2, 3]\n",
    "ticktext = [\"5- anos\", \"5-8 anos\", \"9-11 anos\", \"12+ anos\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d9e1f8d",
   "metadata": {},
   "source": [
    "Das pessoas com menos de 5 anos de escolaridade, 16,4% abandonaram o tratamento;  \n",
    "Com 5 a 8 anos de escolaridade, 21,7%;  \n",
    "Com 9 a 11 anos, 14,9%;  \n",
    "Com mais de 12 anos, 7,8%;  \n",
    "Percebe-se uma discrepância considerável da proporção de abandono entre as categorias. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b8cf111",
   "metadata": {},
   "source": [
    "- #### Tipo de entrada do paciente na unidade de saúde & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "138684ef",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"tratamento\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Caso novo\", \"Retratamento\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a729050",
   "metadata": {},
   "source": [
    "Dentre os casos de pessoas sendo tratadas pela 1ª vez, 13,4% resultaram em abandono. Já dentre as pessoas em retratamento (podendo ser recidiva após cura ou reingresso após abandono), 38% abandonaram.  \n",
    "Para essa variável em específico nota-se uma grande diferença entre os dois possíveis casos."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8af60667",
   "metadata": {},
   "source": [
    "- #### Resultado da radiografia do tórax & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "56ad6117",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"raiox_tora\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1, 2]\n",
    "ticktext = [\"Normal\", \"Suspeito\", \"Não realizado\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0a11ad50",
   "metadata": {},
   "source": [
    "Dos pacientes com resultado normal da radiografia do tórax, 13,2% abandonaram;  \n",
    "Com resultado suspeito (77,3% dos casos), 16% abandonaram;\n",
    "Dos que não realizaram, 21% abandonaram.\n",
    "Observa-se uma tendência maior de abandono em caso de não-realização do raio-x."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0aedcab4",
   "metadata": {},
   "source": [
    "- #### Agravo AIDS associado à tuberculose & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "10451b21",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"agravaids\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Não\", \"Sim\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c67f29a6",
   "metadata": {},
   "source": [
    "Nos casos em que não houve o agravo da tuberculose associado à AIDS por ocasião da notificação, 15,9% abandonaram, enquanto 31,7% abandonaram no caso de haver agravo, ou seja, praticamente o dobro."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6477b8ac",
   "metadata": {},
   "source": [
    "- #### Agravo Alcoolismo associado à tuberculose & Abandono\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4c5954ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"agravalcoo\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Não\", \"Sim\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f23e4c9b",
   "metadata": {},
   "source": [
    "Nos casos em que não houve o agravo no tratamento da tuberculose associado ao alcoolismo, 14,6% abandonaram, enquanto 27% abandonaram no caso de haver agravo, ou seja, também é quase o dobro."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "157a8630",
   "metadata": {},
   "source": [
    "- #### Agravo Diabetes associado à tuberculose & Abandono\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4fb894f6",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"agravdiabe\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Não\", \"Sim\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a052d32",
   "metadata": {},
   "source": [
    "Nos casos em que não houve o agravo no tratamento da tuberculose associado à diabetes por ocasião da notificação, 17,8% abandonaram, enquanto 9,8% abandonaram no caso de haver agravo. Nesse caso, ao contrário dos dois últimos analisados, o agravo relacionado à diabetes resulta em uma chance quase 2x menor de abandono do tratamento da tuberculose. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6dd4c1ea",
   "metadata": {},
   "source": [
    "- #### Agravo Doenças Mentais associado à tuberculose & Abandono\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ecdbab40",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"agravdoenc\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Não\", \"Sim\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2281388",
   "metadata": {},
   "source": [
    "Nos casos em que não houve o agravo no tratamento da tuberculose associado à doenças mentais, 16,9% abandonaram, enquanto 19,6% abandonaram no caso de haver agravo. Nesse caso não há uma diferença notável."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e2261104",
   "metadata": {},
   "source": [
    "- #### Agravo Uso de drogas ilícitas associado à tuberculose & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8d615874",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"agravdroga\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Não\", \"Sim\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e31325d",
   "metadata": {},
   "source": [
    "Nos casos em que não houve o agravo no tratamento da tuberculose associado ao uso de drogas ilícitas, 13,6% abandonaram, enquanto 39,2% abandonaram no caso de haver agravo, ou seja, praticamente o triplo, sendo essa uma diferença bastante relevante."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d37f7d47",
   "metadata": {},
   "source": [
    "- #### Agravo Tabagismo associado à tuberculose & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "93316eb7",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"agravtabac\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Não\", \"Sim\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ea66377",
   "metadata": {},
   "source": [
    "Nos casos em que não houve o agravo no tratamento da tuberculose associado ao tabagismo, 14,3% abandonaram, enquanto 24,9% abandonaram no caso de haver agravo, ou seja, uma chance de 1,75x maior."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b40b1b7a",
   "metadata": {},
   "source": [
    "- #### HIV & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "05df583a",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"hiv\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1, 2]\n",
    "ticktext = [\"Negativo\", \"Positivo\", \"Não realizado\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "224f4639",
   "metadata": {},
   "source": [
    "Nos casos em que os pacientes não realizaram o teste para o HIV, 23,9% abandonaram o tratamento da tuberculose, enquanto 14,4% abandonaram em caso de resultado negativo e 31% abandonaram em caso de resultado positivo ao teste, ou seja, há uma chance consideravelmente maior para esse último caso."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4bd09dd",
   "metadata": {},
   "source": [
    "- #### Tratamento Diretamente Observado (TDO) & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4b98e111",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"tratsup_at\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Não\", \"Sim\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "894c9304",
   "metadata": {},
   "source": [
    "Dentre as pessoas com tuberculose que não receberam TDO, 22,6% abandonaram o tratamento, enquanto 10,4% das que receberam abandonaram. Nota-se que ao haver TDO, a chance do abandono é praticamente 2x menor."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1a9f09dd",
   "metadata": {},
   "source": [
    "- #### Beneficiário de programa de transferência de renda do governo & Abandono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5868f196",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig = px.histogram(df, x=\"benef_gov\", color=\"abandono\",width=400, height=400)\n",
    "\n",
    "tickvals = [0, 1]\n",
    "ticktext = [\"Não\", \"Sim\"]\n",
    "\n",
    "fig.update_xaxes(tickvals=tickvals, ticktext=ticktext)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "77693aa6",
   "metadata": {},
   "source": [
    "17% dos pacientes que são beneficiários de programas de transferência abandonaram o tratamento, enquanto 16,5% dos que não são abandonaram, portanto não se verifica uma diferença relevante."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11a253d9",
   "metadata": {},
   "source": [
    "### Conclusão da análise"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b80987a6",
   "metadata": {},
   "source": [
    "A partir da análise feita, conclui-se que dentre todas as features categóricas do dataset, as seguintes aparentam estar menos relacionadas a uma maior ou menor chance de abandono do tratamento: região de notificação, agravo associado a doenças mentais (agravdoenc) e beneficiário de programa de transferência de renda do governo (benef_gov). Já as mais relacionadas aparentam ser: tipo de entrada na unidade de saúde (tratamento), agravo relacionado ao uso de drogas ilícitas (agravdrog) e TDO (tratsup_at).\n",
    "Podemos analisar o grau de dependência entre as variáveis categóricas e a variável alvo (abandono) usando `mutual_info_score`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "2d3467f1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mutual_info_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>tratamento</th>\n",
       "      <td>0.022286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>agravdroga</th>\n",
       "      <td>0.021751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tratsup_at</th>\n",
       "      <td>0.013449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hiv</th>\n",
       "      <td>0.008493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>agravalcoo</th>\n",
       "      <td>0.007705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>agravtabac</th>\n",
       "      <td>0.006967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cs_escol_n</th>\n",
       "      <td>0.005405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>agravaids</th>\n",
       "      <td>0.004563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cs_sexo</th>\n",
       "      <td>0.003015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>agravdiabe</th>\n",
       "      <td>0.002387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cs_raca</th>\n",
       "      <td>0.002229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>raiox_tora</th>\n",
       "      <td>0.001473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>reg_not</th>\n",
       "      <td>0.000407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>agravdoenc</th>\n",
       "      <td>0.000062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>benef_gov</th>\n",
       "      <td>0.000006</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            mutual_info_score\n",
       "tratamento           0.022286\n",
       "agravdroga           0.021751\n",
       "tratsup_at           0.013449\n",
       "hiv                  0.008493\n",
       "agravalcoo           0.007705\n",
       "agravtabac           0.006967\n",
       "cs_escol_n           0.005405\n",
       "agravaids            0.004563\n",
       "cs_sexo              0.003015\n",
       "agravdiabe           0.002387\n",
       "cs_raca              0.002229\n",
       "raiox_tora           0.001473\n",
       "reg_not              0.000407\n",
       "agravdoenc           0.000062\n",
       "benef_gov            0.000006"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mutual_info_score\n",
    "def cat_mut_inf(series):\n",
    "    return mutual_info_score(series, df['abandono']) \n",
    "df_categ = df.drop(['abandono', 'idade'], axis=1)\n",
    "df_cat = df_categ.apply(cat_mut_inf) \n",
    "df_cat = df_cat.sort_values(ascending=False).to_frame(name='mutual_info_score') \n",
    "df_cat"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "116acf18",
   "metadata": {},
   "source": [
    "Verifica-se que os valores mostrados condizem com a análise feita."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f84c73ba",
   "metadata": {},
   "source": [
    "A partir disso, os testes e treinamentos dos modelos foram feitos no jupyter `modelando.ipynb`."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}