StockPredictor

This is a simple stock predictor object using Python and its Data Science libraries which includes Sklearn, Pandas, Numpy, and yFinance.
It will use Linear Regression or Support Vector Regression to predict the data.

Prerequisites

Python 2.7 minimum, Python 3.x recommended as I created this with Python 3.8 environment.

pip 20.x.x recommended but will work with lower versions.

Python libraries:

numpy
pandas
simplejson
sklearn (install using scikit-learn)
yfinance

Installation

pip install simplejson
pip install -U scikit-learn
pip install yfinance

Usage

import StockPredictor as StockPredictor
stock_predictor = StockPredictor(ticker_symbol, prediction_days, machine_learning_type)
stock_predictor.get_json()

Ticker Symbol is a one to five character symbol that represents a company’s stock name. You must use a NYSE ticker symbol or it will not work.

Prediction Days is the amount of days you would like to predict. I recommend a range from 1-30, no fewer, no more. I would do a hard cap on 30 as well as the prediction will be less accurate.

Machine Learning Type is the option to either use Linear Regression (‘LR’) or Support Vector Regression (‘SVR’).

By default, the Ticker Smybol will be ‘MSFT’, Prediction Days will be ‘10’, and Machine Learning Type will be ‘LR’.

Output

You will get a JSON data type when you use .get_json (Duh!).

It will include the predicted data and the confidence of the prediction.

Here is an example:

{
  "Day 1": 10
  "Day 2": 12
  "Day 3": 13.4
  "Day 4": 12.7
  "Day 5": 11.5
  "Confidence": 0.923322443
}

Contributing

Pull requests are welcome, but it will be unlikely I will change this repo unless it’s a dramatic change. This is to show my understanding of both Python, Object-Oriention, and Data Science.

License

MIT