{"cells":[{"cell_type":"code","execution_count":1,"metadata":{},"outputs":[{"data":{"text/html":" \n "},"metadata":{},"output_type":"display_data"},{"data":{"text/html":" \n "},"metadata":{},"output_type":"display_data"}],"source":"import sys\nsys.path.append(\"../\")\n\nimport settings\n\nimport pycrfsuite\n\nimport json\n\nimport db.db_settings as db_settings\nfrom db.database_connection import DatabaseConnection\n\nfrom Tagging.conllu_generator import ConlluGenerator\nfrom Tagging.crf_data_generator import *\n\nfrom RecipeAnalysis.Recipe import Ingredient\n\nfrom difflib import SequenceMatcher\n\nimport numpy as np\n\nimport plotly.graph_objs as go\nfrom plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot\nfrom plotly.subplots import make_subplots\ninit_notebook_mode(connected=True)\n\nfrom graphviz import Digraph\n\nimport itertools\n\nimport random\n\nimport plotly.io as pio\npio.renderers.default = \"jupyterlab\"\n\nfrom IPython.display import Markdown, HTML, display\n\nfrom copy import deepcopy"},{"cell_type":"code","execution_count":2,"metadata":{},"outputs":[],"source":"import dill\nm_act = dill.load(open(\"m_act.dill\", \"rb\"))\nm_mix = dill.load(open(\"m_mix.dill\", \"rb\"))\nm_base_act = dill.load(open(\"m_base_act.dill\", \"rb\"))\nm_base_mix = dill.load(open(\"m_base_mix.dill\", \"rb\"))\n\n#m_act.apply_threshold(3)\n#m_mix.apply_threshold(3)\n#m_base_act.apply_threshold(5)\n#m_base_mix.apply_threshold(5)\n\n\n#c_act = m_act.get_csr()\n#c_mix = m_mix.get_csr()\n#c_base_act = m_base_act.get_csr()\n#c_base_mix = m_base_mix.get_csr()\n\nc_act = m_act._csr\nc_mix = m_mix._csr\nc_base_act = m_base_act._csr\nc_base_mix = m_base_mix._csr"},{"cell_type":"code","execution_count":3,"metadata":{},"outputs":[],"source":"base_ingredients = m_base_mix.get_labels()\nactions = m_act.get_labels()[0]"},{"cell_type":"markdown","execution_count":null,"metadata":{},"outputs":[],"source":"## get most used ingredients"},{"cell_type":"code","execution_count":18,"metadata":{},"outputs":[],"source":"ings = []\ning_sum = []\n\nfor ing in base_ingredients:\n ings.append(ing)\n ing_sum.append(m_base_mix.get_sum(ing))"},{"cell_type":"code","execution_count":20,"metadata":{},"outputs":[],"source":"ing_sort = np.argsort(-np.array(ing_sum))"},{"cell_type":"code","execution_count":23,"metadata":{},"outputs":[],"source":"w = np.array(ing_sum)[ing_sort[:20]]\ntop_ings = np.array(ings)[ing_sort[:20]]"},{"cell_type":"code","execution_count":24,"metadata":{},"outputs":[{"data":{"text/plain":"array(['salt', 'sugar', 'water', 'butter', 'onion', 'egg', 'flour',\n 'olive oil', 'milk', 'sprinkle', 'pepper', 'cream', 'garlic clove',\n 'cheese', 'tomato', 'dough', 'sauce', 'black pepper', 'roll',\n 'vanilla extract'], dtype='