Details
-
Bug
-
Resolution: Unresolved
-
Major
-
Cheshire-Cat
-
Untriaged
-
Unknown
Description
On both N1QL and analytics, there is a logical bug. Flavor is computed in infer using samples collected during the scan phase. If num_sample_values is 1, then flavor computation breaks down.
On N1QL, for beer sample and num_sample_values = 1, gives flavor of
"Flavor": "`abv` = 7.5, `brewery_id` = \"cigar_city_brewing\", `description` = \"Jai Alai IPA is a monster inter...\", `ibu` = 0, `name` = \"Jai Alai IPA\", `srm` = 0, `type` = \"beer\", `upc` = 0, `updated` = \"2010-07-22 20:00:20\"" |
instead of
type=beer
|
flavor.
On cbas
"Flavor": "`abv` = 5.0, `brewery_id` = \"21st_amendment_brewery_cafe\", `description` = \"Deep black color, toasted black burnt coffee flavors and aroma. Dispensed with Nitrogen through a slow-flow faucet giving it the characteristic cascading effect, resulting in a rich dense creamy head.\", `ibu` = 0.0, `name` = \"563 Stout\", `srm` = 0.0, `type` = \"beer\", `upc` = 0, `updated` = \"2010-07-22 20:00:20\"", |