Ajustado teste F

fcc84468 · mgy20 · f0686af4 · fcc84468 · fcc84468 · fcc84468
Commit fcc84468 authored 1 year ago by mgy20
--- a/Testes_R/.ipynb_checkpoints/TESTE_COHEND-checkpoint.ipynb
+++ b/Testes_R/.ipynb_checkpoints/TESTE_COHEND-checkpoint.ipynb
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "b977bf15-ee73-4730-8651-f8a3caffb393",
+   "metadata": {},
+   "source": [
+    "# Analise estatistica com Teste CohenD\n",
+    "Teste de effect size"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "36e8ccea-ef95-4da0-bca9-45bd2fbe5530",
+   "metadata": {},
+   "source": [
+    "## Instalação dos pacotes\n",
+    "Para fazer o jupyter-lab processar R após a instaçação, dentro do ambiente virtual do python e dentro seu client com sudo `sudo R`, instale os pacotes com o comando a seguir."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "id": "57a4780e-976b-4832-b04a-4680e62ccb5a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#install.packages(c('repr', 'IRdisplay', 'evaluate', 'crayon', 'pbdZMQ', 'devtools', 'uuid', 'digest'))\n",
+    "#devtools::install_github('IRkernel/IRkernel')\n",
+    "#IRkernel::installspec()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "ceab7323-0150-4a5c-8b6f-07464f948b9a",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<table class=\"dataframe\">\n",
+       "<caption>A data.frame: 6 × 10</caption>\n",
+       "<thead>\n",
+       "\t<tr><th></th><th scope=col>ANO_CENSO</th><th scope=col>NUM_COMPUTADOR</th><th scope=col>NUM_COMPUTADOR_ADM</th><th scope=col>NUM_COMPUTADPR_ALUNO</th><th scope=col>NUM_DVD</th><th scope=col>NUM_FUNCIONARIOS</th><th scope=col>NUM_SALAS</th><th scope=col>NUM_SALAS_UTILIZADAS</th><th scope=col>NUM_SOM</th><th scope=col>NUM_TV</th></tr>\n",
+       "\t<tr><th></th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th><th scope=col>&lt;int&gt;</th></tr>\n",
+       "</thead>\n",
+       "<tbody>\n",
+       "\t<tr><th scope=row>1</th><td>2013</td><td>165</td><td>35</td><td>130</td><td>1</td><td>42</td><td> 8</td><td> 8</td><td>1</td><td>3</td></tr>\n",
+       "\t<tr><th scope=row>2</th><td>2013</td><td> 15</td><td> 4</td><td> 11</td><td>5</td><td>16</td><td> 5</td><td> 5</td><td>7</td><td>6</td></tr>\n",
+       "\t<tr><th scope=row>3</th><td>2013</td><td> 25</td><td> 5</td><td> 20</td><td>2</td><td>40</td><td> 8</td><td> 8</td><td>3</td><td>3</td></tr>\n",
+       "\t<tr><th scope=row>4</th><td>2013</td><td> 15</td><td> 8</td><td>  7</td><td>2</td><td>45</td><td>10</td><td>10</td><td>2</td><td>4</td></tr>\n",
+       "\t<tr><th scope=row>5</th><td>2013</td><td> 16</td><td> 8</td><td>  8</td><td>2</td><td>47</td><td>12</td><td>12</td><td>1</td><td>6</td></tr>\n",
+       "\t<tr><th scope=row>6</th><td>2013</td><td> 10</td><td> 7</td><td>  3</td><td>1</td><td>25</td><td> 4</td><td> 4</td><td>0</td><td>1</td></tr>\n",
+       "</tbody>\n",
+       "</table>\n"
+      ],
+      "text/latex": [
+       "A data.frame: 6 × 10\n",
+       "\\begin{tabular}{r|llllllllll}\n",
+       "  & ANO\\_CENSO & NUM\\_COMPUTADOR & NUM\\_COMPUTADOR\\_ADM & NUM\\_COMPUTADPR\\_ALUNO & NUM\\_DVD & NUM\\_FUNCIONARIOS & NUM\\_SALAS & NUM\\_SALAS\\_UTILIZADAS & NUM\\_SOM & NUM\\_TV\\\\\n",
+       "  & <int> & <int> & <int> & <int> & <int> & <int> & <int> & <int> & <int> & <int>\\\\\n",
+       "\\hline\n",
+       "\t1 & 2013 & 165 & 35 & 130 & 1 & 42 &  8 &  8 & 1 & 3\\\\\n",
+       "\t2 & 2013 &  15 &  4 &  11 & 5 & 16 &  5 &  5 & 7 & 6\\\\\n",
+       "\t3 & 2013 &  25 &  5 &  20 & 2 & 40 &  8 &  8 & 3 & 3\\\\\n",
+       "\t4 & 2013 &  15 &  8 &   7 & 2 & 45 & 10 & 10 & 2 & 4\\\\\n",
+       "\t5 & 2013 &  16 &  8 &   8 & 2 & 47 & 12 & 12 & 1 & 6\\\\\n",
+       "\t6 & 2013 &  10 &  7 &   3 & 1 & 25 &  4 &  4 & 0 & 1\\\\\n",
+       "\\end{tabular}\n"
+      ],
+      "text/markdown": [
+       "\n",
+       "A data.frame: 6 × 10\n",
+       "\n",
+       "| <!--/--> | ANO_CENSO &lt;int&gt; | NUM_COMPUTADOR &lt;int&gt; | NUM_COMPUTADOR_ADM &lt;int&gt; | NUM_COMPUTADPR_ALUNO &lt;int&gt; | NUM_DVD &lt;int&gt; | NUM_FUNCIONARIOS &lt;int&gt; | NUM_SALAS &lt;int&gt; | NUM_SALAS_UTILIZADAS &lt;int&gt; | NUM_SOM &lt;int&gt; | NUM_TV &lt;int&gt; |\n",
+       "|---|---|---|---|---|---|---|---|---|---|---|\n",
+       "| 1 | 2013 | 165 | 35 | 130 | 1 | 42 |  8 |  8 | 1 | 3 |\n",
+       "| 2 | 2013 |  15 |  4 |  11 | 5 | 16 |  5 |  5 | 7 | 6 |\n",
+       "| 3 | 2013 |  25 |  5 |  20 | 2 | 40 |  8 |  8 | 3 | 3 |\n",
+       "| 4 | 2013 |  15 |  8 |   7 | 2 | 45 | 10 | 10 | 2 | 4 |\n",
+       "| 5 | 2013 |  16 |  8 |   8 | 2 | 47 | 12 | 12 | 1 | 6 |\n",
+       "| 6 | 2013 |  10 |  7 |   3 | 1 | 25 |  4 |  4 | 0 | 1 |\n",
+       "\n"
+      ],
+      "text/plain": [
+       "  ANO_CENSO NUM_COMPUTADOR NUM_COMPUTADOR_ADM NUM_COMPUTADPR_ALUNO NUM_DVD\n",
+       "1 2013      165            35                 130                  1      \n",
+       "2 2013       15             4                  11                  5      \n",
+       "3 2013       25             5                  20                  2      \n",
+       "4 2013       15             8                   7                  2      \n",
+       "5 2013       16             8                   8                  2      \n",
+       "6 2013       10             7                   3                  1      \n",
+       "  NUM_FUNCIONARIOS NUM_SALAS NUM_SALAS_UTILIZADAS NUM_SOM NUM_TV\n",
+       "1 42                8         8                   1       3     \n",
+       "2 16                5         5                   7       6     \n",
+       "3 40                8         8                   3       3     \n",
+       "4 45               10        10                   2       4     \n",
+       "5 47               12        12                   1       6     \n",
+       "6 25                4         4                   0       1     "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "# Carrega CSV\n",
+    "options(warn = -1)\n",
+    "library(effsize)\n",
+    "df = read.csv(\"../dados/escola-integers-ks.csv\", sep=\"|\")\n",
+    "head(df)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "2c7eca10-6c0b-427d-be50-c4bf3d224f2d",
+   "metadata": {},
+   "source": [
+    "## Limpeza de Outliers\n",
+    "Para limpar dados que se diferenciam drasticamente do resto do conjunto de dados é utilizado a remoção dos outliers."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "ee18ee6d-1dda-4c65-8347-f3b7e980e83c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Function to remove outliers using the IQR method\n",
+    "remove_outliers <- function(data) {\n",
+    "  # Calculate the interquartile range (IQR)\n",
+    "  Q1 <- quantile(data, 0.25)\n",
+    "  Q3 <- quantile(data, 0.75)\n",
+    "  IQR <- Q3 - Q1\n",
+    "  \n",
+    "  # Define the lower and upper bounds\n",
+    "  lower_bound <- Q1 - 1.5 * IQR\n",
+    "  upper_bound <- Q3 + 1.5 * IQR\n",
+    "  \n",
+    "  # Remove outliers\n",
+    "  cleaned_data <- data[data >= lower_bound & data <= upper_bound]\n",
+    "  \n",
+    "  # Return the cleaned data\n",
+    "  return(cleaned_data)\n",
+    "}"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "7ca6c624-62c0-4fce-b3cd-c47d20c0b85e",
+   "metadata": {},
+   "source": [
+    "## Plot de Grafico\n",
+    "Funções para plot de gráficos"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "4c578156-afa1-4432-acb3-d6a143c0b349",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "plot_ecdf <- function(x, x2){\n",
+    "    plot(ecdf(x),\n",
+    "         xlim = range(c(x, x2)), \n",
+    "         col = \"blue\")\n",
+    "    plot(ecdf(x2), \n",
+    "         add = TRUE, \n",
+    "         lty = \"dashed\",\n",
+    "         col = \"red\")\n",
+    "}"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "69c80406-3d20-408d-8fce-e4a7aea65dd4",
+   "metadata": {},
+   "source": [
+    "## Teste Cohen D\n",
+    "Utilizando testes de effect size é possivel determinar a magnitude da relação entre duas amostras\n",
+    "\n",
+    "Primeiro é selecionado duas amostras que foram agrupadas pelos anos do censo.  \n",
+    "Em seguida é feito uma limpeza dos dados fazendo a retirada dos nulos, outliers e os ordenando.  \n",
+    "Finalmente é executado o teste."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "id": "a9629868-ef70-4073-856b-fa3cee413898",
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "[1] -0.03455845\n"
+     ]
+    },
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAACXBIWXMAABJ0AAASdAHeZh94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\nhCHmHljmmQAAAABJRU5ErkJggg==",
+      "text/plain": [
+       "Plot with title “ecdf(x)”"
+      ]
+     },
+     "metadata": {
+      "image/png": {
+       "height": 420,
+       "width": 420
+      }
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "# Dados agrupados por ano\n",
+    "ANO <- df[\"ANO_CENSO\"]\n",
+    "\n",
+    "# Amostras\n",
+    "X <- df[\"NUM_SALAS\"]\n",
+    "Xi <- X[ANO == 2013]\n",
+    "\n",
+    "Y <- df[\"NUM_SALAS\"]\n",
+    "Yi <- Y[ANO == 2014]\n",
+    "\n",
+    "# Limpeza dos dados\n",
+    "data1 <- sort(remove_outliers(na.omit(Xi)))\n",
+    "data2 <- sort(remove_outliers(na.omit(Yi)))\n",
+    "\n",
+    "# Executa teste Cohen D\n",
+    "#print(cohen.d(data1, data2))\n",
+    "result = cohen.d(data1, data2)\n",
+    "print(result$estimate)\n",
+    "\n",
+    "# Plotagem dos dados\n",
+    "plot_ecdf(data1, data2)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "b6c8c14d-5101-4201-b172-28e5b17016df",
+   "metadata": {},
+   "source": [
+    "## Teste CohenD - Anos subsequentes\n",
+    "Execução do teste T entre uma coluna de um ano contra todas as colunas dos anos subsequentes para todos os anos do csv."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 26,
+   "id": "9aa86d14-0e60-4945-9d5f-d058b3c45654",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Colunas do csv de saida\n",
+    "colunas = c(\"coluna1\", \"ano_coluna1\", \"coluna2\", \"ano_coluna2\", \"tamanho_amostra1\", \"estatistica_cohend\")\n",
+    "output_df = data.frame(matrix(ncol = length(colunas), nrow = 0));\n",
+    "\n",
+    "# Remove ANO_CENSO das iteracoes\n",
+    "atributos = names(df)\n",
+    "atributos = atributos[atributos != \"ANO_CENSO\"]\n",
+    "            \n",
+    "# Separa os anos em amostras\n",
+    "ANO <- df[\"ANO_CENSO\"]\n",
+    "\n",
+    "for(ano in sort(unique(df$ANO_CENSO))){\n",
+    "    for(col1 in atributos) {\n",
+    "        for(col2 in atributos) {\n",
+    "\n",
+    "            # Amostra de um ano\n",
+    "            X <- df[col1]\n",
+    "            Xi <- X[ANO == ano]\n",
+    "\n",
+    "            # Amostra do ano seguinte\n",
+    "            Y <- df[col2]\n",
+    "            Yi <- Y[ANO == ano+1]\n",
+    "\n",
+    "            # Remove NaN, outliers e ordena\n",
+    "            data1 <- sort(remove_outliers(na.omit(Xi)))\n",
+    "            data2 <- sort(remove_outliers(na.omit(Yi)))\n",
+    "\n",
+    "            # Pula casos em que não há dados nas amostras\n",
+    "            if(length(data1) == 0 || length(data2) == 0){\n",
+    "                next\n",
+    "            }\n",
+    "\n",
+    "            # Teste Cohen D\n",
+    "            resultado = cohen.d(data1, data2)\n",
+    "\n",
+    "            # Concatena resultados no dataframe\n",
+    "            nova_linha = c(col1, ano, col2, ano+1, length(data1), resultado$estimate)\n",
+    "            output_df = rbind(output_df, nova_linha)\n",
+    "        }\n",
+    "    }\n",
+    "}\n",
+    "output_csv = \"Result_COHEND/COHEND_subsequente.csv\"\n",
+    "colnames(output_df) <- colunas\n",
+    "write.csv(output_df, file = output_csv, row.names = FALSE)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "d52a6568-693a-4d59-b050-2f8d6899433e",
+   "metadata": {},
+   "source": [
+    "## Teste CohenD - Anos acumulados\n",
+    "O mesmo teste anterior, porém com o acumulo da amostra conforme o passar dos anos"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 27,
+   "id": "bd76fa5b-a42c-434a-b3df-83f12f80a262",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Colunas do csv de saida\n",
+    "colunas = c(\"coluna1\", \"ano_coluna1\", \"coluna2\", \"ano_coluna2\", \"tamanho_amostra1\", \"estatistica_cohend\")\n",
+    "output_df = data.frame(matrix(ncol = length(colunas), nrow = 0))\n",
+    "ano_start = min(df$ANO_CENSO)\n",
+    "\n",
+    "# Remove ANO_CENSO das iteracoes\n",
+    "atributos = names(df)\n",
+    "atributos = atributos[atributos != \"ANO_CENSO\"]\n",
+    "            \n",
+    "# Separa os anos em amostras\n",
+    "ANO_COLUMN <- df[\"ANO_CENSO\"]\n",
+    "\n",
+    "for(ano in sort(unique(df$ANO_CENSO))){\n",
+    "    for(col1 in atributos) {\n",
+    "        for(col2 in atributos) {\n",
+    "\n",
+    "            # Amostra acumulada dos anos\n",
+    "            X <- df[col1]\n",
+    "            Xi <- X[ANO_COLUMN >= ano_start & ANO_COLUMN <= ano]\n",
+    "\n",
+    "            # Amostra do ano seguinte\n",
+    "            Y <- df[col2]\n",
+    "            Yi <- Y[ANO_COLUMN == ano+1]\n",
+    "\n",
+    "            # Remove NaN, outliers e ordena\n",
+    "            data1 <- sort(remove_outliers(na.omit(Xi)))\n",
+    "            data2 <- sort(remove_outliers(na.omit(Yi)))\n",
+    "\n",
+    "            # Pula casos em que não há dados nas amostras\n",
+    "            if(length(data1) == 0 || length(data2) == 0){\n",
+    "                next\n",
+    "            }\n",
+    "\n",
+    "            # Teste Cohend\n",
+    "            resultado = cohen.d(data1, data2)\n",
+    "\n",
+    "            # Concatena resultados no dataframe\n",
+    "            nova_linha = c(col1, ano, col2, ano+1, length(data1), resultado$estimate)\n",
+    "            output_df = rbind(output_df, nova_linha)\n",
+    "        }\n",
+    "    }\n",
+    "}\n",
+    "output_csv = \"Result_COHEND/COHEND_acumulado.csv\"\n",
+    "colnames(output_df) <- colunas\n",
+    "write.csv(output_df, file = output_csv, row.names = FALSE)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "R",
+   "language": "R",
+   "name": "ir"
+  },
+  "language_info": {
+   "codemirror_mode": "r",
+   "file_extension": ".r",
+   "mimetype": "text/x-r-source",
+   "name": "R",
+   "pygments_lexer": "r",
+   "version": "4.4.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
+%% Cell type:markdown id:b977bf15-ee73-4730-8651-f8a3caffb393 tags:
+# Analise estatistica com Teste CohenD
+Teste de effect size
+%% Cell type:markdown id:36e8ccea-ef95-4da0-bca9-45bd2fbe5530 tags:
+## Instalação dos pacotes
+Para fazer o jupyter-lab processar R após a instaçação, dentro do ambiente virtual do python e dentro seu client com sudo `sudo R`, instale os pacotes com o comando a seguir.
+%% Cell type:code id:57a4780e-976b-4832-b04a-4680e62ccb5a tags:
+``` R
+#install.packages(c('repr', 'IRdisplay', 'evaluate', 'crayon', 'pbdZMQ', 'devtools', 'uuid', 'digest'))
+#devtools::install_github('IRkernel/IRkernel')
+#IRkernel::installspec()
+```
+%% Cell type:code id:ceab7323-0150-4a5c-8b6f-07464f948b9a tags:
+``` R
+# Carrega CSV
+options(warn = -1)
+library(effsize)
+df = read.csv("../dados/escola-integers-ks.csv", sep="|")
+head(df)
+```
+%% Output
+    A data.frame: 6 × 10
+    | <!--/--> | ANO_CENSO &lt;int&gt; | NUM_COMPUTADOR &lt;int&gt; | NUM_COMPUTADOR_ADM &lt;int&gt; | NUM_COMPUTADPR_ALUNO &lt;int&gt; | NUM_DVD &lt;int&gt; | NUM_FUNCIONARIOS &lt;int&gt; | NUM_SALAS &lt;int&gt; | NUM_SALAS_UTILIZADAS &lt;int&gt; | NUM_SOM &lt;int&gt; | NUM_TV &lt;int&gt; |
+    |---|---|---|---|---|---|---|---|---|---|---|
+    | 1 | 2013 | 165 | 35 | 130 | 1 | 42 |  8 |  8 | 1 | 3 |
+    | 2 | 2013 |  15 |  4 |  11 | 5 | 16 |  5 |  5 | 7 | 6 |
+    | 3 | 2013 |  25 |  5 |  20 | 2 | 40 |  8 |  8 | 3 | 3 |
+    | 4 | 2013 |  15 |  8 |   7 | 2 | 45 | 10 | 10 | 2 | 4 |
+    | 5 | 2013 |  16 |  8 |   8 | 2 | 47 | 12 | 12 | 1 | 6 |
+    | 6 | 2013 |  10 |  7 |   3 | 1 | 25 |  4 |  4 | 0 | 1 |
+    A data.frame: 6 × 10
+    \begin{tabular}{r|llllllllll}
+      & ANO\_CENSO & NUM\_COMPUTADOR & NUM\_COMPUTADOR\_ADM & NUM\_COMPUTADPR\_ALUNO & NUM\_DVD & NUM\_FUNCIONARIOS & NUM\_SALAS & NUM\_SALAS\_UTILIZADAS & NUM\_SOM & NUM\_TV\\
+      & <int> & <int> & <int> & <int> & <int> & <int> & <int> & <int> & <int> & <int>\\
+    \hline
+    	1 & 2013 & 165 & 35 & 130 & 1 & 42 &  8 &  8 & 1 & 3\\
+    	2 & 2013 &  15 &  4 &  11 & 5 & 16 &  5 &  5 & 7 & 6\\
+    	3 & 2013 &  25 &  5 &  20 & 2 & 40 &  8 &  8 & 3 & 3\\
+    	4 & 2013 &  15 &  8 &   7 & 2 & 45 & 10 & 10 & 2 & 4\\
+    	5 & 2013 &  16 &  8 &   8 & 2 & 47 & 12 & 12 & 1 & 6\\
+    	6 & 2013 &  10 &  7 &   3 & 1 & 25 &  4 &  4 & 0 & 1\\
+    \end{tabular}
+%% Cell type:markdown id:2c7eca10-6c0b-427d-be50-c4bf3d224f2d tags:
+## Limpeza de Outliers
+Para limpar dados que se diferenciam drasticamente do resto do conjunto de dados é utilizado a remoção dos outliers.
+%% Cell type:code id:ee18ee6d-1dda-4c65-8347-f3b7e980e83c tags:
+``` R
+# Function to remove outliers using the IQR method
+remove_outliers <- function(data) {
+  # Calculate the interquartile range (IQR)
+  Q1 <- quantile(data, 0.25)
+  Q3 <- quantile(data, 0.75)
+  IQR <- Q3 - Q1
+  # Define the lower and upper bounds
+  lower_bound <- Q1 - 1.5 * IQR
+  upper_bound <- Q3 + 1.5 * IQR
+  # Remove outliers
+  cleaned_data <- data[data >= lower_bound & data <= upper_bound]
+  # Return the cleaned data
+  return(cleaned_data)
+}
+```
+%% Cell type:markdown id:7ca6c624-62c0-4fce-b3cd-c47d20c0b85e tags:
+## Plot de Grafico
+Funções para plot de gráficos
+%% Cell type:code id:4c578156-afa1-4432-acb3-d6a143c0b349 tags:
+``` R
+plot_ecdf <- function(x, x2){
+    plot(ecdf(x),
+         xlim = range(c(x, x2)),
+         col = "blue")
+    plot(ecdf(x2),
+         add = TRUE,
+         lty = "dashed",
+         col = "red")
+}
+```
+%% Cell type:markdown id:69c80406-3d20-408d-8fce-e4a7aea65dd4 tags:
+## Teste Cohen D
+Utilizando testes de effect size é possivel determinar a magnitude da relação entre duas amostras
+Primeiro é selecionado duas amostras que foram agrupadas pelos anos do censo.
+Em seguida é feito uma limpeza dos dados fazendo a retirada dos nulos, outliers e os ordenando.
+Finalmente é executado o teste.
+%% Cell type:code id:a9629868-ef70-4073-856b-fa3cee413898 tags:
+``` R
+# Dados agrupados por ano
+ANO <- df["ANO_CENSO"]
+# Amostras
+X <- df["NUM_SALAS"]
+Xi <- X[ANO == 2013]
+Y <- df["NUM_SALAS"]
+Yi <- Y[ANO == 2014]
+# Limpeza dos dados
+data1 <- sort(remove_outliers(na.omit(Xi)))
+data2 <- sort(remove_outliers(na.omit(Yi)))
+# Executa teste Cohen D
+#print(cohen.d(data1, data2))
+result = cohen.d(data1, data2)
+print(result$estimate)
+# Plotagem dos dados
+plot_ecdf(data1, data2)
+```
+%% Output
+    [1] -0.03455845
+%% Cell type:markdown id:b6c8c14d-5101-4201-b172-28e5b17016df tags:
+## Teste CohenD - Anos subsequentes
+Execução do teste T entre uma coluna de um ano contra todas as colunas dos anos subsequentes para todos os anos do csv.
+%% Cell type:code id:9aa86d14-0e60-4945-9d5f-d058b3c45654 tags:
+``` R
+# Colunas do csv de saida
+colunas = c("coluna1", "ano_coluna1", "coluna2", "ano_coluna2", "tamanho_amostra1", "estatistica_cohend")
+output_df = data.frame(matrix(ncol = length(colunas), nrow = 0));
+# Remove ANO_CENSO das iteracoes
+atributos = names(df)
+atributos = atributos[atributos != "ANO_CENSO"]
+# Separa os anos em amostras
+ANO <- df["ANO_CENSO"]
+for(ano in sort(unique(df$ANO_CENSO))){
+    for(col1 in atributos) {
+        for(col2 in atributos) {
+            # Amostra de um ano
+            X <- df[col1]
+            Xi <- X[ANO == ano]
+            # Amostra do ano seguinte
+            Y <- df[col2]
+            Yi <- Y[ANO == ano+1]
+            # Remove NaN, outliers e ordena
+            data1 <- sort(remove_outliers(na.omit(Xi)))
+            data2 <- sort(remove_outliers(na.omit(Yi)))
+            # Pula casos em que não há dados nas amostras
+            if(length(data1) == 0 || length(data2) == 0){
+                next
+            }
+            # Teste Cohen D
+            resultado = cohen.d(data1, data2)
+            # Concatena resultados no dataframe
+            nova_linha = c(col1, ano, col2, ano+1, length(data1), resultado$estimate)
+            output_df = rbind(output_df, nova_linha)
+        }
+    }
+}
+output_csv = "Result_COHEND/COHEND_subsequente.csv"
+colnames(output_df) <- colunas
+write.csv(output_df, file = output_csv, row.names = FALSE)
+```
+%% Cell type:markdown id:d52a6568-693a-4d59-b050-2f8d6899433e tags:
+## Teste CohenD - Anos acumulados
+O mesmo teste anterior, porém com o acumulo da amostra conforme o passar dos anos
+%% Cell type:code id:bd76fa5b-a42c-434a-b3df-83f12f80a262 tags:
+``` R
+# Colunas do csv de saida
+colunas = c("coluna1", "ano_coluna1", "coluna2", "ano_coluna2", "tamanho_amostra1", "estatistica_cohend")
+output_df = data.frame(matrix(ncol = length(colunas), nrow = 0))
+ano_start = min(df$ANO_CENSO)
+# Remove ANO_CENSO das iteracoes
+atributos = names(df)
+atributos = atributos[atributos != "ANO_CENSO"]
+# Separa os anos em amostras
+ANO_COLUMN <- df["ANO_CENSO"]
+for(ano in sort(unique(df$ANO_CENSO))){
+    for(col1 in atributos) {
+        for(col2 in atributos) {
+            # Amostra acumulada dos anos
+            X <- df[col1]
+            Xi <- X[ANO_COLUMN >= ano_start & ANO_COLUMN <= ano]
+            # Amostra do ano seguinte
+            Y <- df[col2]
+            Yi <- Y[ANO_COLUMN == ano+1]
+            # Remove NaN, outliers e ordena
+            data1 <- sort(remove_outliers(na.omit(Xi)))
+            data2 <- sort(remove_outliers(na.omit(Yi)))
+            # Pula casos em que não há dados nas amostras
+            if(length(data1) == 0 || length(data2) == 0){
+                next
+            }
+            # Teste Cohend
+            resultado = cohen.d(data1, data2)
+            # Concatena resultados no dataframe
+            nova_linha = c(col1, ano, col2, ano+1, length(data1), resultado$estimate)
+            output_df = rbind(output_df, nova_linha)
+        }
+    }
+}
+output_csv = "Result_COHEND/COHEND_acumulado.csv"
+colnames(output_df) <- colunas
+write.csv(output_df, file = output_csv, row.names = FALSE)
+```
--- a/Testes_R/Result_COHEND/.ipynb_checkpoints/COHEND_subsequente-checkpoint.csv
+++ b/Testes_R/Result_COHEND/.ipynb_checkpoints/COHEND_subsequente-checkpoint.csv
--- a/Testes_R/Result_COHEND/COHEND_subsequente.csv
+++ b/Testes_R/Result_COHEND/COHEND_subsequente.csv
--- a/Testes_R/Result_F/.ipynb_checkpoints/F_subsequente-checkpoint.csv
+++ b/Testes_R/Result_F/.ipynb_checkpoints/F_subsequente-checkpoint.csv
-"coluna1","ano_coluna1","coluna2","ano_coluna2","tamanho_amostra1","estatistica_F","p_valor"
+"coluna1","ano_coluna1","coluna2","ano_coluna2","tamanho_amostra1","estatistica_f","p_valor"
 "ANO_CENSO","2007","ANO_CENSO","2008","237387","NaN","NaN"
 "ANO_CENSO","2007","NUM_COMPUTADOR","2008","237387","0","0"
 "ANO_CENSO","2007","NUM_COMPUTADOR_ADM","2008","237387","0","0"

--- a/Testes_R/Result_F/F_acumulado.csv
+++ b/Testes_R/Result_F/F_acumulado.csv
--- a/Testes_R/Result_KS/.ipynb_checkpoints/KS_acumulado-checkpoint.csv
+++ b/Testes_R/Result_KS/.ipynb_checkpoints/KS_acumulado-checkpoint.csv
--- a/Testes_R/Result_KS/KS_acumulado.csv
+++ b/Testes_R/Result_KS/KS_acumulado.csv
--- a/Testes_R/Result_T/.ipynb_checkpoints/T_acumulado-checkpoint.csv
+++ b/Testes_R/Result_T/.ipynb_checkpoints/T_acumulado-checkpoint.csv
--- a/Testes_R/Result_T/T_acumulado.csv
+++ b/Testes_R/Result_T/T_acumulado.csv
--- a/Testes_R/TESTE_COHEND.ipynb
+++ b/Testes_R/TESTE_COHEND.ipynb
--- a/Testes_R/TESTE_F.ipynb
+++ b/Testes_R/TESTE_F.ipynb
@@ -32,7 +32,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 21,
+   "execution_count": 1,
   "id": "ceab7323-0150-4a5c-8b6f-07464f948b9a",
   "metadata": {},
   "outputs": [
@@ -122,7 +122,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 22,
+   "execution_count": 2,
   "id": "ee18ee6d-1dda-4c65-8347-f3b7e980e83c",
   "metadata": {},
   "outputs": [],
@@ -157,7 +157,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 23,
+   "execution_count": 3,
   "id": "4c578156-afa1-4432-acb3-d6a143c0b349",
   "metadata": {},
   "outputs": [],
@@ -262,7 +262,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 25,
+   "execution_count": 4,
   "id": "9aa86d14-0e60-4945-9d5f-d058b3c45654",
   "metadata": {},
   "outputs": [],

--- a/analise_resultados_muriki.ipynb
+++ b/analise_resultados_muriki.ipynb