Skip to content

Instantly share code, notes, and snippets.

@akshaybaweja

akshaybaweja/recipe_analysis.ipynb

Last active April 5, 2021 19:01
Show Gist options
  • Save akshaybaweja/5401170e8bf816f64ad9cfeceead1eb9 to your computer and use it in GitHub Desktop.
Save akshaybaweja/5401170e8bf816f64ad9cfeceead1eb9 to your computer and use it in GitHub Desktop.
Download ZIP

Revisions

  1. akshaybaweja revised this gist Apr 5, 2021. 1 changed file with 906 additions and 83 deletions.
    989 changes: 906 additions & 83 deletions recipe_analysis.ipynb
    View file
    906 additions, 83 deletions not shown because the diff is too large. Please use a local Git client to view these changes.
  2. akshaybaweja revised this gist Apr 5, 2021. 1 changed file with 62 additions and 862 deletions.
    924 changes: 62 additions & 862 deletions recipe_analysis.ipynb
    View file
    62 additions, 862 deletions not shown because the diff is too large. Please use a local Git client to view these changes.
  3. akshaybaweja revised this gist Apr 5, 2021. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion recipe_analysis.ipynb
    View file
    Original file line number Diff line number Diff line change
    @@ -1293,7 +1293,7 @@
    "source": [
    "def WC(values):\n",
    " wc = WordCloud(width = 1920, height = 1080, background_color ='black', stopwords = stopwords, min_font_size = 5,colormap='autumn', mask=transformed_india_mask).generate(values)\n",
    " plt.figure(figsize=(20,11))\n",
    " plt.figure()\n",
    " plt.title(\"TranslatedIngredients\")\n",
    " plt.imshow(wc)\n",
    " plt.axis('off')\n",
  4. akshaybaweja revised this gist Apr 5, 2021. 1 changed file with 0 additions and 324 deletions.
    324 changes: 0 additions & 324 deletions recipe_analysis.ipynb
    View file
    Original file line number Diff line number Diff line change
    @@ -1319,330 +1319,6 @@
    }
    }
    ]
    },
    {
    "cell_type": "markdown",
    "metadata": {
    "id": "p6Jif7cLmwkk"
    },
    "source": [
    "# Machine Learning"
    ]
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "BN-8F28Au5Wu"
    },
    "source": [
    "#data = df.loc[:, 'TranslatedInstructions']\n",
    "data = df.loc[:, 'TranslatedIngredients']"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "markdown",
    "metadata": {
    "id": "iGIlyTL7vsyf"
    },
    "source": [
    "## Training"
    ]
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "k7xSyzx6zvMg"
    },
    "source": [
    "from gensim.utils import simple_preprocess\n",
    "from gensim.models.doc2vec import Doc2Vec, TaggedDocument\n",
    "import re\n",
    "\n",
    "tagged_data = []\n",
    "for i, line in enumerate(data):\n",
    " tokens = simple_preprocess(line)\n",
    " tagged_data.append(TaggedDocument(tokens, [i]))"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "b-ZzI72szwT9"
    },
    "source": [
    "# Generally it is better to use less epochs for larger datasets\n",
    "# This is a small dataset so 30 works well. See article for recommendations\n",
    "\n",
    "model = Doc2Vec(vector_size=50, min_count=2, epochs=30)"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "mjFLXEhmz0bs"
    },
    "source": [
    "model.build_vocab(tagged_data)"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "colab": {
    "base_uri": "https://localhost:8080/"
    },
    "id": "W8Wz0Ipbz27D",
    "outputId": "a52a64e3-ab2f-4ebd-9701-7c02b69b1abb"
    },
    "source": [
    "%time model.train(tagged_data, total_examples=model.corpus_count, epochs=model.epochs)"
    ],
    "execution_count": null,
    "outputs": [
    {
    "output_type": "stream",
    "text": [
    "CPU times: user 38 s, sys: 5.24 s, total: 43.2 s\n",
    "Wall time: 27 s\n"
    ],
    "name": "stdout"
    }
    ]
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "EVSY_dkjz54s"
    },
    "source": [
    "model.save('/content/drive/MyDrive/Colab Notebooks/indianFood.model')"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "QWkH3-Mb0ADh"
    },
    "source": [
    "model = Doc2Vec.load(\"/content/drive/MyDrive/Colab Notebooks/indianFood.model\")"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "U46Ot24y0At5"
    },
    "source": [
    "features_extracted = []\n",
    "for i, review in enumerate(data):\n",
    " features_extracted.append(model.docvecs[i])"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "5tRA4nyS0E5t"
    },
    "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "ss = StandardScaler()\n",
    "scaled = ss.fit_transform(features_extracted)"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "2OVY2JPM0Has"
    },
    "source": [
    "from sklearn.cluster import KMeans\n",
    "\n",
    "kmeans = KMeans(n_clusters=20)\n",
    "clusters = kmeans.fit_predict(scaled)"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "1xVjcMFA0bFt"
    },
    "source": [
    "from sklearn.neighbors import NearestNeighbors\n",
    "\n",
    "neighbors = NearestNeighbors(n_neighbors=5, metric='cosine').fit(scaled)\n",
    "_, closest = neighbors.kneighbors(kmeans.cluster_centers_)"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "markdown",
    "metadata": {
    "id": "TOfLk79E57lu"
    },
    "source": [
    "## UMAP"
    ]
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "cK0UmLea59Eu"
    },
    "source": [
    "import umap"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "colab": {
    "base_uri": "https://localhost:8080/"
    },
    "id": "FqN9OKjf5-uP",
    "outputId": "b4cb0c7c-c66d-4e12-e36e-b8cbb809fb1e"
    },
    "source": [
    "%time embedding = umap.UMAP().fit_transform(scaled)"
    ],
    "execution_count": null,
    "outputs": [
    {
    "output_type": "stream",
    "text": [
    "CPU times: user 42.4 s, sys: 477 ms, total: 42.9 s\n",
    "Wall time: 28.3 s\n"
    ],
    "name": "stdout"
    }
    ]
    },
    {
    "cell_type": "code",
    "metadata": {
    "id": "uuzQCKlv6Ad5"
    },
    "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "scaler = MinMaxScaler()\n",
    "scaler.fit(embedding)\n",
    "embedding_scaled = scaler.transform(embedding)"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "colab": {
    "background_save": true
    },
    "id": "MA-oB3tJ6Iop"
    },
    "source": [
    "# Now we put all our data into one list\n",
    "lookup = []\n",
    "\n",
    "for index, umap_pos in enumerate(embedding_scaled):\n",
    " lookup.append({\n",
    " 'cluster_num': float(clusters[index]),\n",
    " 'umap_pos': umap_pos.tolist(),\n",
    " 'name': df.iloc[index]['TranslatedRecipeName'],\n",
    " 'ingredients': df.iloc[index]['TranslatedIngredients'],\n",
    " 'preptime': int(df.iloc[index]['PrepTimeInMins']),\n",
    " 'cooktime': int(df.iloc[index]['CookTimeInMins']),\n",
    " 'totaltime': int(df.iloc[index]['TotalTimeInMins']),\n",
    " 'servings': int(df.iloc[index]['Servings']),\n",
    " 'cuisine': df.iloc[index]['Cuisine'],\n",
    " 'course': df.iloc[index]['Course'],\n",
    " 'diet': df.iloc[index]['Diet'],\n",
    " 'instructions': df.iloc[index]['TranslatedInstructions'],\n",
    " 'image': df.iloc[index]['image'],\n",
    " 'url': df.iloc[index]['URL']\n",
    " })"
    ],
    "execution_count": null,
    "outputs": []
    },
    {
    "cell_type": "code",
    "metadata": {
    "colab": {
    "background_save": true
    },
    "id": "bstVdpbJ6M5B",
    "outputId": "761473e4-743f-4b29-d21a-ca10e9630310"
    },
    "source": [
    "# This is what one entry in the list looks like\n",
    "lookup[1]"
    ],
    "execution_count": null,
    "outputs": [
    {
    "output_type": "execute_result",
    "data": {
    "text/plain": [
    "{'cluster_num': 12.0,\n",
    " 'cooktime': 10,\n",
    " 'course': 'Main Course',\n",
    " 'cuisine': 'South Indian Recipes',\n",
    " 'diet': 'Vegetarian',\n",
    " 'image': 'https://www.archanaskitchen.com/images/archanaskitchen/1-Author/b.yojana-gmail.com/Spicy_Thakkali_Rice_Tomato_Pulihora-1_edited.jpg',\n",
    " 'ingredients': '2-1 / 2 cups rice - cooked, 3 tomatoes, 3 teaspoons BC Belle Bhat powder, salt - as per taste, 1 teaspoon chana dal, 1/2 teaspoon cumin seeds, 1 teaspoon white urad dal, 1/2 Teaspoon mustard, 1 green chilli, 1 dry red chilli, 2 teaspoon cashew - or peanuts, 1-1 / 2 tablespoon oil - 1/2 teaspoon asafoetida ',\n",
    " 'instructions': 'To make tomato puliogere, first cut the tomatoes. Now put in a mixer grinder and puree it. Now heat oil in a pan. After the oil is hot, add chana dal, urad dal, cashew nuts and let it cook for 10 to 20 seconds. After 10 to 20 seconds, add cumin seeds, mustard seeds, green chillies, dry red chillies and curry leaves. After 30 seconds, add tomato puree to it and mix. Add BC Belle Bhat powder, salt and mix it. Allow to cook for 7 to 8 minutes and then turn off the gas. Take it out in a bowl, add cooked rice and mix it. Serve hot. Serve tomato puliogre with tomato cucumber raita and papad for dinner. ',\n",
    " 'name': 'Spicy Tomato Rice (Recipe) ',\n",
    " 'preptime': 5,\n",
    " 'servings': 3,\n",
    " 'totaltime': 15,\n",
    " 'umap_pos': [0.20701216161251068, 0.297481507062912],\n",
    " 'url': 'http://www.archanaskitchen.com/spicy-tomato-rice-recipe-in-hindi'}"
    ]
    },
    "metadata": {
    "tags": []
    },
    "execution_count": 0
    }
    ]
    },
    {
    "cell_type": "code",
    "metadata": {
    "colab": {
    "background_save": true
    },
    "id": "JHUsWA8vuFGs"
    },
    "source": [
    "import json\n",
    "\n",
    "with open('/content/drive/MyDrive/Colab Notebooks/Recipe Classification/data.json', 'w') as outfile:\n",
    " json.dump(lookup, outfile)"
    ],
    "execution_count": null,
    "outputs": []
    }
    ]
    }
  5. akshaybaweja revised this gist Apr 5, 2021. 1 changed file with 12 additions and 1 deletion.
    13 changes: 12 additions & 1 deletion recipe_analysis.ipynb
    View file
    Original file line number Diff line number Diff line change
    @@ -6,14 +6,25 @@
    "name": "recipe_analysis.ipynb",
    "provenance": [],
    "collapsed_sections": [],
    "toc_visible": true
    "toc_visible": true,
    "include_colab_link": true
    },
    "kernelspec": {
    "display_name": "Python 3",
    "name": "python3"
    }
    },
    "cells": [
    {
    "cell_type": "markdown",
    "metadata": {
    "id": "view-in-github",
    "colab_type": "text"
    },
    "source": [
    "<a href=\"https://colab.research.google.com/gist/akshaybaweja/5401170e8bf816f64ad9cfeceead1eb9/recipe_analysis.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
    ]
    },
    {
    "cell_type": "markdown",
    "metadata": {
  6. akshaybaweja created this gist Apr 5, 2021.
    1,637 changes: 1,637 additions & 0 deletions recipe_analysis.ipynb
    View file
    1,637 additions, 0 deletions not shown because the diff is too large. Please use a local Git client to view these changes.