| ALS cold start user/item prediction strategy | 3 sec | Passed |
| ALS should not introduce unnecessary shuffle | 0.67 sec | Passed |
| CheckedCast | 0.46 sec | Passed |
| CholeskySolver | 1 ms | Passed |
| LocalIndexEncoder | 3 ms | Passed |
| RatingBlockBuilder | 2 ms | Passed |
| SPARK-18268: ALS with empty RDD should fail with better message | 23 ms | Passed |
| UncompressedInBlock | 19 ms | Passed |
| als partitioner is a projection | 3 ms | Passed |
| als with large number of iterations | 7.1 sec | Passed |
| approximate rank-1 matrix | 4.2 sec | Passed |
| approximate rank-2 matrix | 4.3 sec | Passed |
| case insensitive cold start param value | 6.6 sec | Passed |
| different block settings | 7.9 sec | Passed |
| exact rank-1 matrix | 4.9 sec | Passed |
| implicit feedback | 2.2 sec | Passed |
| implicit feedback regression | 3.2 sec | Passed |
| input type validation | 46 sec | Passed |
| more blocks than ratings | 2 sec | Passed |
| nonnegative constraint | 0.81 sec | Passed |
| normal equation construction | 2 ms | Passed |
| partitioner in returned factors | 0.44 sec | Passed |
| read/write | 1.7 sec | Passed |
| recommendForAllItems with k <, = and > num_users | 2 sec | Passed |
| recommendForAllUsers with k <, = and > num_items | 2.6 sec | Passed |
| recommendForItemSubset with k <, = and > num_users | 2.1 sec | Passed |
| recommendForUserSubset with k <, = and > num_items | 2.3 sec | Passed |
| subset recommendations eliminate duplicate ids, returns same results as unique ids | 2.1 sec | Passed |
| subset recommendations on full input dataset equivalent to recommendForAll | 1.5 sec | Passed |
| using generic ID types | 1.7 sec | Passed |
![[Jenkins]](/jenkins/static/86f3da3d/images/jenkins-header-logo-v2.svg){kind=logo}
