Hands-On Machine Learning for Algorithmic Trading

Building a price/earnings time series

In total, the nine years of filing history provide us with over 28,000 numerical values. We can select a useful field, such as Earnings per Diluted Share (EPS), that we can combine with market data to calculate the popular Price/Earnings (P/E) valuation ratio.

We do need to take into account, however, that Apple split its stock 7:1 on June 4, 2014, and Adjusted Earnings per Share before the split to make earnings comparable, as illustrated in the following code block:

field = 'EarningsPerShareDiluted'stock_split = 7split_date = pd.to_datetime('20140604')# Filter by tag; keep only values measuring 1 quartereps = aapl_nums[(aapl_nums.tag == 'EarningsPerShareDiluted') & (aapl_nums.qtrs == 1)].drop('tag', ...