Example of a custom factor using fundamental data¶

# Imports needed to create and run pipeline
from quantopian.pipeline import Pipeline, CustomFactor  
from quantopian.research import run_pipeline  

# Import any pipeline data we want to use. These are most common.
from quantopian.pipeline.data.builtin import USEquityPricing
from quantopian.pipeline.data import Fundamentals  # Morningstar
from quantopian.pipeline.data import factset   # Factset

# Import any built in filters we want to use
from quantopian.pipeline.filters import StaticAssets

# Import these just in case
import numpy as np
import pandas as pd

# Check the numpy version. Later versions have a nicer 'unique' method which can be applied over an axis
# Looks like numpy isn't new enough so we'll need to map the function across columns manually
np.__version__

# Make a list of securities to test with
my_securities = StaticAssets(symbols(['T', 'VZ', 'FTR']))
my_securities

# Define out custom factor
# It will return the percent of change and the days since the change
# Note that the days is NOT the days since the data was reported 
class DataChange(CustomFactor):
    window_length=90
    outputs= ["pct_change", "days_since_change"]
              
    def compute(self, today, assets, out, data):

        def column_change(column):
                # The numpy 'unique' returns the unique values and the 
                # indexes of the first occurances of those values
                value_at_changes, index_at_changes = np.unique(column, return_index=True)

                # Note that the 'unique' method sorts the results by low to high values
                # We really want them in the original time-sequenced order
                # So sort the index and fetch the associated values from the original data
                index_at_changes = sorted(index_at_changes)
                value_at_changes = column[index_at_changes]
                
                if len(index_at_changes) == 1:
                    # The 'unique' method only found a single value - no data change within the window
                    change = 0.0
                    days_since_change = np.nan
                else:
                    # There was a change. We just want the last change so check [-1] and [-2]
                    change = (value_at_changes[-1] / value_at_changes[-2]) - 1.0
                    days_since_change = self.window_length - index_at_changes[-1] 

                return change, days_since_change
        
        # Apply the above function across each column and output the values
        out.pct_change[:], out.days_since_change[:] = np.apply_along_axis(column_change, 0, data)

# As a check, lets make a naive change factor
# It just checks if todays value is different fom yesterdays value
class DataChange_naive(CustomFactor):

    def compute(self, today, assets, out, data):
        out[:] = (data[-1] / data[-2]) - 1.0

def make_pipeline():
    cash_change = DataChange(inputs=[Fundamentals.cash_and_cash_equivalents],
            window_length = 90,
            mask = my_securities)
    
    naive_cash_change = DataChange_naive(inputs=[Fundamentals.cash_and_cash_equivalents],
            window_length = 90,
            mask = my_securities)
    
    quarterly_sales_change = DataChange(inputs=[factset.Fundamentals.sales_qf],
            window_length = 90,
            mask = my_securities)

    return Pipeline(
        columns={
            'cash_change_pct' : cash_change.pct_change,
            'cash_change_days' : cash_change.days_since_change,
            'cash_change_test' : naive_cash_change,
            'quarterly_sales_change_pct' : quarterly_sales_change.pct_change,
        },
        screen = my_securities
    )
df = run_pipeline(make_pipeline(), '2018-3-2', '2018-3-2')
#df['total_revenue_ttm_asof_date'] = df['total_revenue_ttm_asof_date'].astype('datetime64[ns]')
df