Red de bicibletas - NetworkX.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Estructura de la red\n",
    "La red de bicicletas puede ser considerada como una estructura de procesamiento de informacion donde los nodos son las unidades de procesamiento y el camino que los vincula, la conexión.\n",
    "\n",
    "Estructura de la red:\n",
    "    \n",
    "    1) NODOS: Estaciones\n",
    "    2) CONEXION (edge): Camino | Recorrido | Calles\n",
    "    3) PESOS (intensidad de la conexion): cantidad de usuarios? | tiempo de uso? | distancia en km?\n",
    "    4) DIRECCIONALIDAD: Dirigido o No dirigido?  Podria pensarse como dirigido o dirigido reciproco con self-loops\n",
    "    5) TAMAÑO\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Metricas de la red y preguntas relevantes\n",
    "\n",
    "1) Por cuantos nodos esta compuesta la red y cuantos edges los vinculan?\n",
    "\n",
    "2) Como es la direccionalidad de la red: \n",
    "    2.1_Las bicicletas suelen dejarse en la misma estacion de la que se retiran?(self-loop)\n",
    "    2.2_Cuales son las estaciones que mas conexiones tienen entre si?\n",
    "\n",
    "3) Degree centrality distribution de los nodos:\n",
    "    3.1_Cual es la estacion con mayor cantidad de conexiones?\n",
    "    \n",
    "4) Betweenness centrality distribution de los nodos:\n",
    "    4.1_Hay alguna estacion que permita clusterizar recorridos?\n",
    "    \n",
    "5) Cual es el grado promedio de la red? El promedio de relaciones que existe entre los nodos de la red (<k> = L/N)\n",
    "    \n",
    "6) Cual es el flujo maximo de pasajeros que podria soportar la red de bicicletas?\n",
    "\n",
    "7) Cual es el camino mas corto entre n y k estaciones?\n",
    "\n",
    "8) Diametro de la red. Cual es la distancia promedio que se recorre en la red de bicicletas de caba?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Librerias que vamos a usar\n",
    "import matplotlib.pyplot as plt\n",
    "import networkx as nx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1: Hay x nodos conectados por n edges\n",
    "len(B.nodes())\n",
    "len(B.edges())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Plot the degree distribution de la red de bicicletas\n",
    "plt.hist(list(nx.degree_centrality(G).values()))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Plot the betweennes centrality distribution de la red de bicicletas\n",
    "plt.hist(list(nx.betweenness_centrality(G).values()))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Representar los ejes con un matrixplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cual es la estacion con mayor degree centrality?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Se pueden recomendar conexiones entre estaciones?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ESTRUCTURANDO LA RED DE BICICLETAS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cargamos datasets de recorridos\n",
    "df15 = pd.read_csv('data/bici15_cdn.csv')\n",
    "df16 = pd.read_csv('data/bici16_cdn.csv')\n",
    "df17 = pd.read_csv('data/bici17_cdn.csv')\n",
    "df18 = pd.read_csv('data/bici18_cdn.csv')\n",
    "df19 = pd.read_csv('data/bici19_cdn.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import networkx as nx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "networkx.classes.digraph.DiGraph"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Instanciamos un grafo direccionado\n",
    "D = nx.DiGraph()\n",
    "type(D)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Esto me parece que seria pensando a las personas como nodo\n",
    "# Funcion para crear edges en una red de transporte\n",
    "\n",
    "def vinculos_e_nodos(df, columna1, columna2):\n",
    "    '''\n",
    "    Devuelve los extremos conectados por un vinculo (edge)\n",
    "    '''\n",
    "    \n",
    "    origen = df[columna1]\n",
    "    destino = df[columna2]\n",
    "    \n",
    "    vinculados = zip(origen,destino)\n",
    "    \n",
    "    return list(vinculados)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "edges = vinculos_e_nodos(df18, 'bici_estacion_origen', 'bici_estacion_destino')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creamos una lista de edges (estacion_n, estacion_k)\n",
    "D.add_edges_from(edges)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "56940"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(D.edges())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "18658"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(D.nodes())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "# A partir de aca seria pensando las estaciones como nodo\n",
    "df_graf = df18.groupby(['bici_nombre_estacion_origen','bici_nombre_estacion_destino']).size().reset_index(name='recorridos')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_graf.columns = ['origen','destino','recorridos']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "grafo = nx.from_pandas_edgelist(df_graf,'origen','destino','recorridos')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6,6))\n",
    "pos = nx.spring_layout(grafo)\n",
    "nx.draw_networkx_nodes(grafo, pos, alpha=0.2, node_size=50)\n",
    "nx.draw_networkx_edges(grafo, pos, style='solid', aplha=0.1)\n",
    "plt.axis('off')\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "import geopandas as gpd\n",
    "import mplleaflet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Shape de estaciones descargados de DataBA\n",
    "estaciones = gpd.read_file('data/estaciones_de_bicicletas.shp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "estaciones = estaciones.to_crs(epsg='4326')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "estaciones = nodos[['NOMBRE','DIRE_NORM', 'geometry']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [],
   "source": [
    "nodos = estaciones[['NOMBRE','DIRE_NORM']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [],
   "source": [
    "nodos['x'], nodos['y'] = estaciones.geometry.x, estaciones.geometry.y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_graf2 = pd.merge(df_graf,nodos,left_on='origen',right_on='NOMBRE' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_graf3 = pd.merge(df_graf2,nodos,left_on='destino',right_on='NOMBRE' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['origen', 'destino', 'recorridos', 'NOMBRE_x', 'DIRE_NORM_x', 'x_x',\n",
       "       'y_x', 'NOMBRE_y', 'DIRE_NORM_y', 'x_y', 'y_y'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_graf3.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_graf3.columns = ['origen', 'destino', 'recorridos', 'nom_or', 'dirnom_or', 'lon_or',\n",
    "       'lat_or', 'nom_de', 'dirnom_de', 'lon_de', 'lat_de']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_graf4 = df_graf3[['origen', 'dirnom_or', 'lon_or', 'lat_or', 'destino', 'dirnom_de', 'lon_de', 'lat_de', 'recorridos']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creamos una lista de adyacencias> un diccionario por nodo (key) y una lista de nodos con los que conecta como value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}