Skip to main content

Forecasting tools

Project description

anticipy

AnticiPy is a tool to generate forecasts for time series. It takes a pandas Series or DataFrame as input, and returns a DataFrame with the forecasted values for a given period of time.

Features:

  • Simple interface. Start forecasting with a single function call on a pandas DataFrame.
  • Model selection. If you provide different multiple models (e.g. linear, sigmoidal, exponential), the tool will compare them and choose the best fit for your data.
  • Trend and seasonality. Support for weekly and monthly seasonality, among other types.
  • Calendar events. Provide lists of special dates, such as holiday seasons or bank holidays, to improve model performance.
  • Data cleaning. The library has tools to identify and remove outliers, and to detect and handle step changes in the data.

It is straightforward to generate a simple linear model with the tool - just call forecast.run_forecast(my_dataframe):

   import pandas as pd, numpy as np
   from anticipy import forecast, forecast_models
   df = pd.DataFrame({'y': np.arange(0., 5)}, index=pd.date_range('2018-01-01', periods=5, freq='D'))
   df_forecast = forecast.run_forecast(df, extrapolate_years=1)
   print df_forecast.head(12)

Output:

   .        date   model             y  is_actuals
   0  2018-01-01       y  0.000000e+00        True
   1  2018-01-02       y  1.000000e+00        True
   2  2018-01-03       y  2.000000e+00        True
   3  2018-01-04       y  3.000000e+00        True
   4  2018-01-05       y  4.000000e+00        True
   5  2018-01-01  linear  5.551115e-17       False
   6  2018-01-02  linear  1.000000e+00       False
   7  2018-01-03  linear  2.000000e+00       False
   8  2018-01-04  linear  3.000000e+00       False
   9  2018-01-05  linear  4.000000e+00       False
   10 2018-01-06  linear  5.000000e+00       False
   11 2018-01-07  linear  6.000000e+00       False

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

anticipy-0.0.2.tar.gz (54.6 kB view hashes)

Uploaded Source

Built Distribution

anticipy-0.0.2-py2-none-any.whl (60.1 kB view hashes)

Uploaded Python 2

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page