1.analisis-tamanos-de-los-picos.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Variación del tamaño de los picos de *Geospiza scandens*\n",
    "\n",
    "En este notebook se analiza la variación de la longitud y la altura picos de las aves durante los años 1943 a 2012, y se realiza una prueba de hipótesis estadística para afirmar si efectivamente ocurrió una variación o se trata de una coincidencia en los datos. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p align=\"center\">\n",
    "  <img width=\"700\" src=\"https://www.notion.so/image/https%3A%2F%2Fs3-us-west-2.amazonaws.com%2Fsecure.notion-static.com%2F81d21818-0ee3-402d-9f62-aaeb292876df%2Fpicos.png?table=block&id=f5396e96-0f64-4089-a328-e0c697b4bc40&width=2130&userId=5d54f20c-b387-4e7e-b2a5-d88e235ada88&cache=v2\">\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np # \n",
    "import pandas as pd # \n",
    "import matplotlib.pyplot as plt \n",
    "import seaborn as sns \n",
    "\n",
    "np.random.seed(42)\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "sns.set() \n",
    "\n",
    "# Importar las funciones matematicas definidas para el analisis\n",
    "%run funciones-matematicas.ipynb\n",
    "\n",
    "# Ajustar tamaño de las figuras en matplotlib\n",
    "plt.rcParams['figure.figsize'] = [10, 8]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['band', 'species', 'Beak length, mm', 'Beak depth, mm'], dtype='object')\n",
      "Index(['band', 'species', 'blength', 'bdepth'], dtype='object')\n"
     ]
    }
   ],
   "source": [
    "# Lectura y preparación de los datos\n",
    "data_1975 = pd.read_csv('./datasets/finch_beaks_1975.csv')\n",
    "data_2012 = pd.read_csv('./datasets/finch_beaks_2012.csv')\n",
    "\n",
    "# Verificar nombres de columnas\n",
    "print(data_1975.columns)\n",
    "print(data_2012.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Igualar nombres de columnas\n",
    "data_2012.columns = data_1975.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Añadir años\n",
    "data_1975['year'] = 1975\n",
    "data_2012['year'] = 2012"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.concat([data_1975, data_2012], axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "band               0\n",
       "species            0\n",
       "Beak length, mm    0\n",
       "Beak depth, mm     0\n",
       "year               0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verificar si existen datos nulos\n",
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# exportar dataframe limpio\n",
    "data.to_csv('./datasets/data_finch_beaks.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Análisis exploratorio "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Geospiza scandens\n",
    "\n",
    "scandens = data[data['species']== 'scandens']\n",
    "\n",
    "_ = sns.swarmplot(x='year', y='Beak depth, mm', data=scandens)\n",
    "_ = plt.ylabel('Tamaño del pico (mm)')\n",
    "_ = plt.xlabel('Año')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Esta gráfica da la impresión de que el tamaño del pico promedio aumento ligeramente en el año 2012, sin embargo, para tener una visualización más precisa, se utilizará la función de distribución acumulativa empírica (ecdf)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Seleccionar del dataframe scandens los valores del tamaño de los picos (bd) en 1975 y 2012\n",
    "bd_1975 = scandens[scandens['year']== 1975]['Beak depth, mm']\n",
    "bd_2012 = scandens[scandens['year']== 2012]['Beak depth, mm']\n",
    "\n",
    "x_1975, y_1975 = ecdf(bd_1975)\n",
    "x_2012, y_2012 = ecdf(bd_2012)\n",
    "\n",
    "# Graficar las ECDF\n",
    "_ = plt.plot(x_1975, y_1975, marker='.', linestyle='none')\n",
    "_ = plt.plot(x_2012, y_2012, marker='.', linestyle='none')\n",
    "\n",
    "# Ajustes del margen\n",
    "plt.margins(0.02)\n",
    "\n",
    "# Ajustes de los ejes y leyendas\n",
    "_ = plt.xlabel('Tamaño del pico (mm)')\n",
    "_ = plt.ylabel('ECDF')\n",
    "_ = plt.legend(('1975', '2012'), loc='lower right')\n",
    "\n",
    "# Mostrar \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Las diferencias son mucho más claras en la ECDF, el promedio es mayor en los datos de 2012 y la varianza también parece aumentar, por lo que se puede concluir que efectivamente los tamaños de los picos han aumentado."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Estimación de parámetros: ¿Qué tan grande es el tamaño del pico en promedio?\n",
    "\n",
    "A continuación se estimará la diferencia promedio de los picos entre 2012 y 1975. Se utilizará la técnica de remuestreo Bootstrapping para presentar los resultados dentro del 95% del intervalo de confianza."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Diferencia de promedios: 0.2262204724409429 mm\n",
      "95% intervalo de confianza = [0.05633521 0.39190544] mm\n"
     ]
    }
   ],
   "source": [
    "# Calcular la diferencia de las medias muestrales\n",
    "mean_diff = np.mean(bd_2012) - np.mean(bd_1975)\n",
    "\n",
    "# Obtener los bootstrap replicates del promedio \n",
    "bs_replicates_1975 = draw_bs_reps(bd_1975, np.mean, size=10000)\n",
    "bs_replicates_2012 = draw_bs_reps(bd_2012, np.mean, size=10000)\n",
    "\n",
    "# Calcular la diferencia de los bootstrap replicates \n",
    "bs_diff_replicates = bs_replicates_2012 - bs_replicates_1975\n",
    "\n",
    "# Calcular el 95% del intervalo de confianza\n",
    "conf_int = np.percentile(bs_diff_replicates, [2.5, 97.5])\n",
    "\n",
    "# Print the results\n",
    "print('Diferencia de promedios:', mean_diff, 'mm')\n",
    "print('95% intervalo de confianza =', conf_int, 'mm')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "El intervalo de confianza nos dice que en el caso de que se pudieran reunir nuevas muestras, existe un 95% de probabilidad de que el valor real de la diferencia de los promedios se encuentre entre 0,056 y 0,392 mm."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test de hipótesis\n",
    "\n",
    "El gráfico de la ECDF y la determinación del intervalo de confianza dejan bastante claro que los picos de G. *scandens* en Daphne Major han aumentado su tamaño. Pero, ¿es posible que este efecto se deba sólo al azar? Para evaluar esta posibilidad, se establecerá la hipótesis de que los tamaños de los picos promedio son iguales en 1975 y 2012, y se calculará la probabilidad de obtener la diferencia observada de 0.226 mm. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `Hipótesis nula (H0):` El tamaño de los picos son iguales en 1975 y 2012. La diferencia observada fue producto del azar. \n",
    "\n",
    "- `Test estadístico`: Diferencia entre los promedios.\n",
    "\n",
    "- `P-value`: La probabilidad de obtener valores iguales o mayores a la diferencia de promedios observada. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "p = 0.0035\n"
     ]
    }
   ],
   "source": [
    "# Calcular el promedio del dataset combinado\n",
    "combined_mean = np.mean(np.concatenate((bd_1975, bd_2012)))\n",
    "\n",
    "# Desplazar las muestras para hacer que tengan igual promedio\n",
    "bd_1975_shifted = bd_1975 - np.mean(bd_1975) + combined_mean\n",
    "bd_2012_shifted = bd_2012 - np.mean(bd_2012) + combined_mean\n",
    "\n",
    "# Obtener bootstrap replicates\n",
    "bs_replicates_1975 = draw_bs_reps(bd_1975_shifted, np.mean, 10000)\n",
    "bs_replicates_2012 = draw_bs_reps(bd_2012_shifted, np.mean, 10000)\n",
    "\n",
    "# Calcular la diferencia entre los bootstrap replicates de 2012 y 1975\n",
    "bs_diff_replicates = bs_replicates_2012 - bs_replicates_1975\n",
    "\n",
    "# Calcular p value\n",
    "p = np.sum(bs_diff_replicates >= mean_diff) / len(bs_diff_replicates)\n",
    "\n",
    "print('p =', p)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Este bajo valor de p-value conduce a rechazar la hipótesis nula inicial, ya que esto significa que la probabilidad de que la diferencia de promedios observada sea producto del azar es de 3.5%, por lo tanto se confirma que el tamaño del pico aumentó con el tiempo."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}