Untitled1.ipynb

Untitled1.ipynb

{
  "cells": [
    {
      "metadata": {
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:00.157685Z",
          "end_time": "2022-05-21T18:11:00.225531Z"
        },
        "trusted": true
      },
      "id": "7d2bd7d3",
      "cell_type": "code",
      "source": "%load_ext autoreload\n%autoreload 2",
      "execution_count": 1,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:00.234415Z",
          "end_time": "2022-05-21T18:11:04.401052Z"
        },
        "trusted": true
      },
      "id": "bf8e0c91",
      "cell_type": "code",
      "source": "from spacy.lang.es.examples import sentences\n\nsentences[1].lower()",
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 2,
          "data": {
            "text/plain": "'los coches autónomos delegan la responsabilidad del seguro en sus fabricantes.'"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:04.415042Z",
          "end_time": "2022-05-21T18:11:05.478399Z"
        },
        "trusted": true
      },
      "id": "1f3ffad0",
      "cell_type": "code",
      "source": "import spacy\n\nnlp = spacy.load(\"es_core_news_sm\")\ndoc = nlp(sentences[1].lower())",
      "execution_count": 3,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:05.488575Z",
          "end_time": "2022-05-21T18:11:05.558617Z"
        },
        "trusted": true
      },
      "id": "b8928bc7",
      "cell_type": "code",
      "source": "for token in doc:\n    print(token.text, token.pos_, token.morph.to_dict())",
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": "los DET {'Definite': 'Def', 'Gender': 'Masc', 'Number': 'Plur', 'PronType': 'Art'}\ncoches NOUN {'Gender': 'Masc', 'Number': 'Plur'}\nautónomos ADJ {'Gender': 'Masc', 'Number': 'Plur'}\ndelegan VERB {'Mood': 'Ind', 'Number': 'Plur', 'Person': '3', 'Tense': 'Pres', 'VerbForm': 'Fin'}\nla DET {'Definite': 'Def', 'Gender': 'Fem', 'Number': 'Sing', 'PronType': 'Art'}\nresponsabilidad NOUN {'Gender': 'Fem', 'Number': 'Sing'}\ndel ADP {'Definite': 'Def', 'Gender': 'Masc', 'Number': 'Sing', 'PronType': 'Art'}\nseguro NOUN {'Gender': 'Masc', 'Number': 'Sing'}\nen ADP {}\nsus DET {'Number': 'Plur', 'Person': '3', 'Poss': 'Yes', 'PronType': 'Prs'}\nfabricantes NOUN {'Number': 'Plur'}\n. PUNCT {'PunctType': 'Peri'}\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:05.583355Z",
          "end_time": "2022-05-21T18:11:05.662552Z"
        }
      },
      "id": "9580ff0b",
      "cell_type": "code",
      "source": "nouns = [token for token in doc if token.pos_ == 'NOUN']\nnouns",
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 5,
          "data": {
            "text/plain": "[coches, responsabilidad, seguro, fabricantes]"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:05.691098Z",
          "end_time": "2022-05-21T18:11:05.825684Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "nouns_dict = [token.morph.to_dict() for token in nouns ]",
      "execution_count": 6,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:05.833655Z",
          "end_time": "2022-05-21T18:11:05.947749Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "nouns_dict",
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 7,
          "data": {
            "text/plain": "[{'Gender': 'Masc', 'Number': 'Plur'},\n {'Gender': 'Fem', 'Number': 'Sing'},\n {'Gender': 'Masc', 'Number': 'Sing'},\n {'Number': 'Plur'}]"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:05.962569Z",
          "end_time": "2022-05-21T18:11:06.040909Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "genders = [nouns_dict[i]['Gender'] for i in range(len(nouns_dict)) if 'Gender' in nouns_dict[i].keys()]",
      "execution_count": 8,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "start_time": "2022-05-21T18:11:06.044794Z",
          "end_time": "2022-05-21T18:11:06.146692Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "genders.count('Masc'), genders.count('Fem')",
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 9,
          "data": {
            "text/plain": "(2, 1)"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "",
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "_draft": {
      "nbviewer_url": "https://gist.github.com/61e29fb558a4ff309040562f6cdd54d2"
    },
    "gist": {
      "id": "61e29fb558a4ff309040562f6cdd54d2",
      "data": {
        "description": "Untitled1.ipynb",
        "public": true
      }
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3 (ipykernel)",
      "language": "python"
    },
    "language_info": {
      "name": "python",
      "version": "3.10.4",
      "mimetype": "text/x-python",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "pygments_lexer": "ipython3",
      "nbconvert_exporter": "python",
      "file_extension": ".py"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}