Skip to content

Latest commit

 

History

History
88 lines (68 loc) · 2.78 KB

README.md

File metadata and controls

88 lines (68 loc) · 2.78 KB

Financial Anomaly Detection

A Toolkit to perform anomaly detection on time-series data of any stock that can be download via yahoo finance API. You can download the stock data and perform anomaly detection using a range of algorithms

Installation

For installing the pre-release version use -

pip install https://github.com/AchintyaX/fad/releases/download/v0.1.1/fad-0.1.1-py3-none-any.whl

Usage

The package has 2 primary methods -

  1. fetch_data - Used for fetching stock data using yahoo finance API and transforming it for anamoly detection :
In [1]: from fad import FinancialAnomalyDetector

In [2]: fad = FinancialAnomalyDetector()

In [3]: start = '2020-01-10'

In [4]: end = '2022-09-10'

In [5]: data = fad.fetch_data('SBIN.NS', start, end)

In [6]: data
Out[6]:
                  Open        High  ...   Adj Close      Volume
2020-01-10  331.000000  337.950012  ...  324.151825  42377838.0
2020-01-11  331.000000  337.950012  ...  324.151825  42377838.0
2020-01-12  331.000000  337.950012  ...  324.151825  42377838.0
2020-01-13  334.000000  335.450012  ...  322.688354  23615129.0
2020-01-14  329.799988  331.700012  ...  320.005402  26296117.0
...                ...         ...  ...         ...         ...
2022-09-06  538.000000  542.700012  ...  537.799988   8657868.0
2022-09-07  534.400024  537.450012  ...  532.849976   8445359.0
2022-09-08  536.000000  546.299988  ...  544.650024  12240707.0
2022-09-09  549.650024  557.000000  ...  553.349976  18585432.0
2022-09-10  549.650024  557.000000  ...  553.349976  18585432.0

[975 rows x 6 columns]
  1. anomaly_detection - Used to find anomalies in the data returning the dataframe along with plotting them in the time series. We have a series algorithms available for usage -
    • seasonal anomaly detection
    • min cluster anomaly detection
    • outlier anomaly detection
    • level shift anomaly detection
    • volatile shift anomaly detection
      You can read more about them here
In [7]: fad.anomaly_detect(data)
Out[7]: 
    2020-01-10    False
    2020-01-11    False
    2020-01-12    False
    2020-01-13    False
    2020-01-14    False
                ...  
    2022-09-06    False
    2022-09-07    False
    2022-09-08    False
    2022-09-09    False
    2022-09-10    False
    Freq: D, Length: 975, dtype: bool

For Development

  1. Clone the project
  2. pip install --upgrade pip
  3. pip install poetry
  4. poetry install

project setup is complete for contributions

For bugs and feature requests please raise an issue on the github repo
For contributions please raise a Pull Request

Credits

  1. yfinance
  2. ADTK