Quantopian Research¶
The Quantopian research platform is a hosted research environment. Our aim is to make it simple to get started on your research without a lot of setup. Libraries are pre-installed and data is made readily accessible.
How do I use this?¶
The Quantopian research platform is build using IPython, an open source project that provides an interactive notebook with executable cells. Each cell can be executed individually, but they all share the same variables (more technically the same namespace).
To execute a cell, click on the cell and press shift + enter.¶
Resarching Investment Algorithms¶
We've added the following functions to optimize the platform for researching and developing investment algorithms.
- get_pricing()
- symbols()
- local_csv()
- get_backtest()
Query for Pricing Data¶
The get_pricing function provides access to 12 years of US Equity pricing data: the same data used by the Quantopian backtester.
get_pricing returns a pandas object. This could be a panel, dataframe or series depending on the input values.
get_pricing('AAPL') #returns a pandas dataframe when run for 1 security
Customize your Queries¶
By default, get_pricing returns OHLCV data at daily frequency for all of 2013. You can change the default behavior by passing the following parameters:
- Specify start and end dates by passing values to
start_dateandend_date. Dates can be passed as pythondatetimeobjects, pandasTimestamps, or strings that can be parsed by pandas. - Narrow the set of returned fields by passing strings or lists of strings as
fields. Valid fields are:priceclose_price(an alias of price)open_pricehighlowvolume
- Access minutely or daily data by setting frequency to
frequency = dailyorfrequency='minute'.
aapl_minute_closes = get_pricing(
'AAPL',
fields='close_price', #modify to price, open_price, high, low or volume to change the field
start_date='2014-01-01', #customize your pricing date range
end_date = '2014-07-01',
frequency='minute', #change to daily for daily pricing
)
# matplotlib is installed for easy plotting
aapl_minute_closes.plot()
Convert Tickers to Symbols¶
By default symbols returns the security object for a ticker symbol. Specify a ticker symbol, or list of symbols, as a string and get a list of security objects back.
- Use
symbol_reference_dateto identify which date you want the symbol back for a particular ticker symbol. - Specify how you would like missing results to be handeled with
handle_missing
symbols('AAPL')
Upload your own CSVs¶
If you want to work with your own data in our environment, you can upload a csv to your data directory and load the file with local_csv. By default, local_csv returns a DataFrame with the data from the loaded file.
You can customize the resulting DataFrame by passing the following parameters:
- Automatically convert strings to security objects by indicating a
symbol_column - Automatically parse dates as Datetime object by indicating a
date_column - Specify what timezone to use with
timezone - Set the date column of your choosing to be the index with
use_date_column_as_index
# Apple_Rev is a CSV from Quandl showing the quarterly revenue for Apple.
# It is pre-loaded in your data folder.
AAPLRev = local_csv('AAPL_Rev.csv', date_column = 'Date', use_date_column_as_index = True)
AAPLRev
# Plot with matplotlib
AAPLRev.plot()
Access Backtest Results (Alpha)¶
The get_backtest function provides programmatic access to the results of backtests run on the Quantopian platform. It takes a single parameter, the ID of a backtest for which results are desired.
You can find the ID of a backtest in the URL of its full results page, which will be of the form:
https://www.quantopian.com/algorithms/<algorithm_id>/<backtest_id>.
result = get_backtest('536a6d181a4f090716c383b7')
The BacktestResult Object¶
The BacktestResult object returned by get_backtest contains all the information stored by Quantopian about the performance of a given backtest run, as well as some additional metadata about the backtest.
Get_backtest returns a backtest results object that includes 11 different data sets. If you type results.<tab> you will see a list of all the things that are available.
result.
Because there are so many fields on BacktestResult, the class provides helper attributes for easy reference.
scalarscontains a list of all the scalar attributes ofBacktestResult.framescontains a list of all theDataFrameattributes ofBacktestResult.attrscontains a list off all the public attributes ofBacktestResult.
print "Scalars:"
print result.scalars
print ""
print "Frames:"
print result.frames
print ""
print "All Attributes:"
print result.attrs
Preview Contents¶
For easy inspection of the contents of frame contents, BacktestResult provides a preview method that returns a transposed version of the first few columns of a frame.
result.preview('risk')
Get Interactive Documentation¶
View documentation for any callable python object with the ? operator.
get_pricing?
import pandas as pd
pd.DataFrame?