Hi folks, just wanna share a few sql questions I had on a recent Amazon Data Scientist phone interview\n\nthere are 2 tables: STREAMS and TITLES\n```\nSTREAMS table\nUSER_ID\nTITLE_ID\nSTREAM_TIMESTART\nSTREAM_MINUTES\nKey for this table is (USER_ID,TITLE_ID,STREAM_TIMESTART)\n\nTITLES table\n\nTITLE_ID\nGENRE\nNAME\nLAUNCH_DATE\n\nKey for this table is TITLE_ID\n```\n\n1. Find the number of times of each movie (title) was streamed and the total minutes streamed and the movie\'s genre in the month of December 2018. \n\tMy solution: Straightforward ```join``` and ```group by```\n2. Find the name and title id of the first movie streamed by each user each day.\n\tMy solution: Window function parition by ```USER_ID``` and ```date(STREAM_TIMESTART)``` and order by ```STREAM_TIMESTART``` here, taking ```row_number()==1```. \n3. Find the total minutes streamed for each movie within the first 28 days since the movie\'s launch date.\n\tMy solution: I did a ```join``` with a filter ```LAUNCH_DATE + INTERVAL 28 DAYS >= date(STREAM_TIMESTART)``` here but the interviewer did not seemed to like/understand this approach and kept asking me for an alternative approach. I guess he found it unintuitive? Anyone else got another idea please?\n\t

Hi folks, just wanna share a few sql questions I had on a recent Amazon Data Scientist phone interview\n\nthere are 2 tables: STREAMS and TITLES\n\nSTREAMS tabl