TSAP is a python package that provides tools for time series analysis in financial data.
Given input of a stock price series, the system will fit time series models, estimate the parameters and do statistical inference. With the identified model, the system predict the future price and assess the prediction accuracy. We can further consider trading strategy and option pricing. Moreover, given the input of multiple stock prices, the system can implement clustering and build a reduced order model for price prediction.
- Download TSAP package from GitHub:
git clone https://github.com/APC524/tsap.git - Add the folder tsap into your Python search path.
TSAP package provides six Python classes.
- AR: the autoregressive model to fit the imput stock price series, computing the log-likelihood and the gradient.
- MR: the moving average model.
- Solver: estimate the model parameters given the model class and the optimization method.
- OptionPricing: calculate the option price given the underlying stock.
- Cluster: impelement the clustering of multiple stock price series.
- Reduction: build a reduced order model for price prediction.
Following is the high-level program structure figure.
- The Project Report explains the detail of the whole project.
- The User Manual gives a brief introduction of the functionality of the package.
- The user can also generate Doxygen HTML and LaTeX manuals with
Doxyfile,
using the command
doxygen Doxyfile. - In demo folder there are several examples showing how to use the package.
This is the course project of APC524/MAE560 Software Engineering for Scientific Computing (Fall 2016) in Princeton University. The project members are Wenyan Gong, Zongxi Li, Cong Ma, Qingcan Wang, Zhuoran Yang and Hao Zhang. We would appreciate Professor Stone and Assistant Instructor Jeffry and Bernat for their guidance and help.