Given a table of daily experiment results, how would you compute cumulative conversion rate and a 7-day rolling average in SQL?
Use the provided daily results table. Treat a day with zero exposures or missing conversions as having no daily conversion rate, while cumulative totals should continue across all dates.
experiment_date, exposures, conversions, daily_conversion_rate, cumulative_conversion_rate, and rolling_7_day_avg.| Column | Type | Description |
|---|---|---|
| experiment_datePK | DATE | Date of the experiment result |
| exposures | INT | Number of users exposed on the date |
| conversions | INT | Number of exposed users who converted |