building-energy-forecastor 0.0.1

Pour l'instant fait pas grand chose

Forecastor for Buildings' Consumption

Go check ce lien pour rédiger le README: Github-flavored Markdown

1. Choses à faire

• FAUT MODIFIER LE SETUP.PY? VOILA VOILA
• Ajouter des features
  • Gérer les exceptions un peu partout pour que ce soit user friendly *_* (simley avec des étoiles dans les yeux)
• Ajouter un modele (voir 2 ou 3 soyons fous)
• comprendre plus encore le sujet mdr

2. Features' List for preprocessing data

• day_of_week(date_serie): Input: serie of dates ; Output: serie of corresponding week days (['Monday', 'Tuesday', ...])
• cycle_time(df): From the 3rd competitor of the Forecast challenge by Schneider Electric ; turns time in a continous element (no break between 23h59 and 00h01) by applying cosinus and sinus functions
• add_weather(df, weather_df): From the 3rd competitor of the Forecast challenge by Schneider Electric ; add the weather data to the training dataset (df here) merging the two dataframes on the 'Timestamp'

3. Requirements for dataframes

• The column containing dates must be labeled 'Timestamp'

4. Model functions

• building_regressor(): returns a linear regressor from Scikit-learn
• building_train(reg, X, y): train the regressor with X the data and Y the targeted values
• building_prediction(reg, X): returns a dataframe showing the prediction of the regressor reg given the data X

5. To install the package

• Command: python -m pip install --index-url https://test.pypi.org/simple/ --no-deps building_energy_forecastor