/TimeSeriesAnalysis

Framework for testing, validating and brainstorming

Primary LanguagePython

TimeSeriesAnalysis

Framework for testing, validating and brainstorming

Dependencies:

  • Numpy
  • Pandas
  • Statsmodels
  • matplotlib
  • pywt

General way of running the code

  • python main.py + -d *STOCK_NAME* + -m *MODEL_NAME*

To transform time series data

  • On command line, run python main.py -d GOOG -m ARIMA
  • GOOG is the name of the csv data in Data directory; ARIMA is the name of the Model class in Model directory.

To perform prediction

  • On command line, run python main.py -d GOOG -t wavelet
  • wavelet is the name of the transformation class in Transformations directory.

To add a new model

  • Create a new model class in Models as a Python file

To add a new transformation

  • Create a new Transformation class in Transformations as a Python file

To add a new time series data

  • Add the csv file into the Data directory