Список проголосовавших за комментарий "ortega, я поменял на shift(-1), вот результат новый <br/> <br/> по позициям юридических лиц по фьючерсу новатэк<br/> <a class="imgpreview" href="/uploads/2024/images/22/65/29/2024/04/02/ff2e0b.png"><img src="/uploads/2024/images/22/65/29/2024/04/02/2f762c.webp"/><br/> <br/> </a>по позициям физичеких лиц по фьючерсу Новатэк<br/> <br/> <a class="imgpreview" href="/uploads/2024/images/22/65/29/2024/04/02/5f6812.png"><img src="/uploads/2024/images/22/65/29/2024/04/02/3fc60e.webp"/><br/> </a><br/> вот код который я использовал:<br/> <p># Calculate daily returns<br/> df['Daily Returns'] = df['price'].pct_change()</p><p># Shift the daily returns by one to apply them to the next day<br/> df['Shifted Daily Returns'] = df['Daily Returns'].shift(-1)</p><p># Strategy returns based on the signal and the shifted daily returns<br/> # When Signal is 1 (buy), we use the shifted daily return as is (benefit from an increase in price next day).<br/> # When Signal is -1 (sell), we invert the shifted daily return (benefit from a decrease in price next day).<br/> # Neutral days have no impact (returns are 0).<br/> df['Strategy Returns'] = np.where(df['Signal'] == 1, df['Shifted Daily Returns'],<br/> np.where(df['Signal'] == -1, -df['Shifted Daily Returns'], 0))</p><p># Calculate cumulative returns by compounding the strategy returns<br/> df['Cumulative Returns'] = (1 + df['Strategy Returns']).cumprod() — 1</p><p># Plot the cumulative returns<br/> fig, ax = plt.subplots(figsize=(12, 8))<br/> ax.plot(df['date'], df['Cumulative Returns'], color='blue')<br/> ax.set_ylabel('Cumulative Returns')<br/> ax.set_title('Strategy Cumulative Returns Over Time')<br/> plt.show()</p><br/> <br/> <br/> "