Stock price extractor and signal generator using Exponential Moving Averages (EMA) at different timeframes:
- 4H -> four hours
- 1D -> one day
- 1W -> and one week EMAs can be plotted at different lengths.
By default it implements an EMA multi-timeframe strategy of crossovers. The strategy considers a BUY signal if the EMA of length 3 at 1D crossed the EMA of length 15 at 1D.
This is done looking at the previous day's EMA 3 and 15 in order to compare. To complete the signal, also EMA of length 5 at 1W needs to be bigger that EMA of length 15 at 1W. The conditions for the BUY signal look like this:
- EMA(3) at 1D > EMA(15) at 1D
- EMA(5) at 1W > EMA(15) at 1W
The reserve conditions are true for SELL signals:
- EMA(3) at 1D < EMA(15) at 1D
- EMA(5) at 1W < EMA(15) at 1W
The plots use japanese candlesticks to draw prices at every time intervals. Each candlestick drawn provides information about 'open', 'close', 'high', and 'low'. Their color represents direction, if it is green signifies that price closed higher than it opened, and if red, the price ended up lower than the open.
In the plots, the EMAs are plotted as lines. The red color coresponds to the smaller length EMAS i.e. EMA (3), EMA(5) for both 1D and 1W. The green color corresponds to the larger length EMAs i.e EMA(15) for both 1D and 1W.
Volume is also included per day. Red color means price closed lower an that day and green color means price went up in that candle.
They offer real-time or historical data on financial assets that are currently being traded in the markets. These APIs usually offer prices of public stocks, ETFs, ETNs.
These data can be used for generating technical indicators which are the foundation to build trading strategies and monitor the market.
Use of the following APIs and where they can be used:
- Yahoo Finance
- Google Finance in Google Sheets
- IEX Cloud
- AlphaVantage
- World trading data
- Other APIs (Polygon.io, Intrinio, Quandl)
This is currently used. There’s a new python module yfinance that wraps the new Yahoo Finance API. To install it:
pip install yfinanceDocumentation for the API can be found here.
To succesfully run this repository, you will need:
- Python >= 3.9
- Miniconda3/Conda
To install it locally:
git clone https://github.com/danielboloc/Supplypipe.git
cd Supplypipe
conda env create --name <env-name> --file environment.yml
conda activate <env-name>
python setup.py --build
python setup.py --installTo pull it form Docker Hub here:
docker pull dboloc/supplypipe:v1.0.0To build the image locally:
cd Supplypipe
export DOCKER_BUILDKIT=1
docker build -t supplypipe:v1.0.0 .All parameters can be seen with:
$> python SupplyPipe.py --help
Usage: Supplypipe.py [OPTIONS]
Options:
--only-stock To only look at the 'stocks' section in config
--on-demand TEXT You can put the ticker symbol, e.g MSFT or multiple 'MSFT,
AAPL'
--intervals TEXT Three intervals for multi-timeframe analysis
--start TEXT Start date of data retrieval. The analysis will be
performed in this interval also
--end TEXT End date of data retrieval
--help Show this message and exit.
Using --on-demand will not take into account the strategy/ies logic although it will plot all components. This parameter is just to see the price action of the stock/s.
Using --only-stock parameter will only consider the securities in the stocks section under SECTORS in config/default.conf.
Time intervals can be specified for the data download. By default is '60m 1d 1wk'. This will download resampled data in the case of one day and week, but for 4H we will have to resample.
You can specify the --start and --end which will limit the days to download the data i.e. could be 2022-07-01 to 2022-07-22, and the plots will only this time window. The max days to display is 729.
There is a SECTORS section in config/default.conf which displays all sectors and has one stock/ETF as defalut. New sectors and known stocks separated by comma in their respective sector.
cd Supplypipe
python -m unittest discoverAnd the tests through Docker:
docker run \
--rm \
--entrypoint /bin/bash dboloc/supplypipe:v1.0.0 \
-c "python -m unittest discover"Use AAPL on demand.
Locally:
python Supplypipe.py --on-demand AAPL
Docker:
docker run \
--mount "type=bind,source=$(pwd)/analysis,destination=/supplypipe/analysis"\
dboloc/supplypipe:v1.0.0 --on-demand AAPLRun in for all securities in the [SECTORS][stocks] config.
Locally:
python Supplypipe.py --only-stock
Docker:
docker run \
--mount "type=bind,source=$(pwd)/analysis,destination=/supplypipe/analysis"\
dboloc/supplypipe:v1.0.0 --only-stockRun in for all sectors in [SECTORS] config.
Locally:
python Supplypipe.py
Docker:
docker run \
--mount "type=bind,source=$(pwd)/analysis,destination=/supplypipe/analysis"\
dboloc/supplypipe:v1.0.0Running this package you will get one or multiple plots depending on the arguments passed. The next sections are generated for AAPL:
- first section: 4H timeframe with EMA(3) and EMA(15) at 4H. The candles are very clear and the EMA crossover is also clear (
redline is abovegreen).
- second section: 1D timeframe with EMA(3) and EMA(15) at 1D. The candles are more abundant, but the time window is bigger.
- thirs section: 1W timeframe with EMA(5) and EMA(15) at 1W. The candles are very clear, with the same time window as 1D.
- fourth section: volume bars for 1D timeframe.
The information contained herein is not intended to be a source of advice or investment analysis with respect to the material presented, and the information and/or code, documents contained in this repository DO NOT constitute investment advice. The ideas and strategies should never be used without first assessing your own personal and financial situation, or without consulting a financial professional. This code is intended to be used and must be used for informational purposes only. I'm not a broker/dealer nor an investment advisor.