Python module to get stock data/cryptocurrencies from the Alpha Vantage API
Alpha Vantage delivers a free API for real time financial data and most used finance indicators in a simple json or pandas format. This module implements a python interface to the free API provided by Alpha Vantage (http://www.alphavantage.co/). It requires a free API, that can be requested on http://www.alphavantage.co/support/#api-key. You can have a look at all the api calls available in their documentation http://www.alphavantage.co/documentation
- From version 1.9.0 onwards, the urllib was substituted by pythons request library that is thread safe. If you have any error, post an issue.
- From version 1.8.0 onwards, the column names of the data frames have changed, they are now exactly what alphavantage gives back in their json response. You can see the examples in better detail in the following git repo: https://github.com/RomelTorres/av_example
- From version 1.6.0, pandas was taken out as a hard dependency.
To install the package use:
pip install alpha_vantageOr install with pandas support, simply install pandas too:
pip install alpha_vantage, pandasIf you want to install from source, then use:
git clone https://github.com/RomelTorres/alpha_vantage.git
pip install -e alpha_vantageTo get data in a python, simply import the library and call the object with your api key and get ready for some awesome free realtime finance data. Your api key may also be stored in the environment variable ALPHAVANTAGE_API_KEY.
from alpha_vantage.timeseries import TimeSeries
ts = TimeSeries(key='YOUR_API_KEY')
# Get json object with the intraday data and another with the call's metadata
data, meta_data = ts.get_intraday('GOOGL')Internally there is a retries counter, that can be used to minimize connection errors (in case that the api is not able to respond in time), the default is set to 5 but can be increased or decreased whenever needed.
ts = TimeSeries(key='YOUR_API_KEY',retries='YOUR_RETRIES')The library supports giving its results as json dictionaries (default), pandas dataframe (if installed) or csv, simply pass the parameter output_format='pandas' to change the format of the output for all the api calls in the given class. Please note that some API calls do not support the csv format (namely ForeignExchange, SectorPerformances and TechIndicators) because the API endpoint does not support the format on their calls either.
ts = TimeSeries(key='YOUR_API_KEY',output_format='pandas')The pandas data frame given by the call, can have either a date string indexing or an integer indexing (by default the indexing is 'data'), depending on your needs, you can use both.
# For the default date string index behavior
ts = TimeSeries(key='YOUR_API_KEY',output_format='pandas', indexing_type='date')
# For the default integer index behavior
ts = TimeSeries(key='YOUR_API_KEY',output_format='pandas', indexing_type='integer')The data frame structure is given by the call on alpha vantage rest API. The column names of the data frames are the ones given by their data structure. For example, the following call:
from alpha_vantage.timeseries import TimeSeries
from pprint import pprint
ts = TimeSeries(key='YOUR_API_KEY', output_format='pandas')
data, meta_data = ts.get_intraday(symbol='MSFT',interval='1min', outputsize='full')
pprint(data.head(2))Would result on:
The headers from the data are specified from Alpha Vantage (in previous versions, the numbers in the headers were removed, but long term is better to have the data exactly as Alpha Vantage produces it.)
Using pandas support we can plot the intra-minute value for 'MSFT' stock quite easily:
from alpha_vantage.timeseries import TimeSeries
import matplotlib.pyplot as plt
ts = TimeSeries(key='YOUR_API_KEY', output_format='pandas')
data, meta_data = ts.get_intraday(symbol='MSFT',interval='1min', outputsize='full')
data['4. close'].plot()
plt.title('Intraday Times Series for the MSFT stock (1 min)')
plt.show()Giving us as output:
The same way we can get pandas to plot technical indicators like Bolliger Bands®
from alpha_vantage.techindicators import TechIndicators
import matplotlib.pyplot as plt
ti = TechIndicators(key='YOUR_API_KEY', output_format='pandas')
data, meta_data = ti.get_bbands(symbol='MSFT', interval='60min', time_period=60)
data.plot()
plt.title('BBbands indicator for MSFT stock (60 min)')
plt.show()Giving us as output:
We can also plot sector performance just as easy:
from alpha_vantage.sectorperformance import SectorPerformances
import matplotlib.pyplot as plt
sp = SectorPerformances(key='YOUR_API_KEY', output_format='pandas')
data, meta_data = sp.get_sector()
data['Rank A: Real-Time Performance'].plot(kind='bar')
plt.title('Real Time Performance (%) per Sector')
plt.tight_layout()
plt.grid()
plt.show()Giving us as output:
We can also plot crypto currencies prices like BTC:
from alpha_vantage.cryptocurrencies import CryptoCurrencies
import matplotlib.pyplot as plt
cc = CryptoCurrencies(key='YOUR_API_KEY', output_format='pandas')
data, meta_data = cc.get_digital_currency_intraday(symbol='BTC', market='CNY')
data['1b. price (USD)'].plot()
plt.tight_layout()
plt.title('Intraday value for bitcoin (BTC)')
plt.grid()
plt.show()Giving us as output:
The foreign exchange is just metadata, thus only available as json format (using the 'csv' or 'pandas' format will raise an Error)
import alpha_vantage.foreignexchange import ForeignExchange
from pprint import pprint
cc = ForeignExchange(key='YOUR_API_KEY')
# There is no metadata in this call
data, _ = cc.get_currency_exchange_rate(from_currency='BTC',to_currency='USD')
pprint(data)Giving us as output:
{
'1. From_Currency Code': 'BTC',
'2. From_Currency Name': 'Bitcoin',
'3. To_Currency Code': 'USD',
'4. To_Currency Name': 'United States Dollar',
'5. Exchange Rate': '5566.80500105',
'6. Last Refreshed': '2017-10-15 15:13:08',
'7. Time Zone': 'UTC'
}
I have added a repository with examples in a python notebook to better see the usage of the library: https://github.com/RomelTorres/av_example
In order to run the tests you have to first export your API key so that the test can use it to run, also the test require pandas, mock and nose.
export API_KEY=YOUR_API_KEY
cd alpha_vantage
nosetestsThe code documentation can be found at https://alpha-vantage.readthedocs.io/en/latest/
Contributing is always welcome, since sometimes I am busy. Just contact me on how best you can contribute.
- The integration tests are not being run at the moment within travis, gotta fix them to run.
- Add test for csv calls as well.
- Add tests for incompatible parameter raise errors.
If you like or use this project, consider showing your support by staring it.
🇻🇪-🇩🇪