dylan created CARBONDATA-1487:
--------------------------------- Summary: problem with the conversion data , using the carbondata struct type Key: CARBONDATA-1487 URL: https://issues.apache.org/jira/browse/CARBONDATA-1487 Project: CarbonData Issue Type: Bug Components: data-load, data-query, spark-integration Affects Versions: 1.1.0 Environment: spark1.6 Reporter: dylan i'm use spark1.6 and carbondata1.1.0. i use struct type,and i find the data is wrong when i load data into carbondata. my test data int csv like this : device,source d803c517c971,NULL d803c517c972,NULL d803c517c973,"{""lat"":null,""lon"":null,""province"":null}" d803c517c974,NULL d803c517c975,NULL d803c517c976,NULL d803c517c977,NULL d803c517c978,"{""lat"":""22.696963"",""lon"":""113.948252"",""province"":""cn2""}" d803c517c979,NULL d803c517c980,NULL d803c517c981,NULL d803c517c982,NULL d803c517c983,NULL d803c517c984,NULL d803c517c985,NULL d803c517c986,"{""lat"":""22.542006"",""lon"":""113.954106"",""province"":""cn2""}" d803c517c987,NULL d803c517c988,NULL d803c517c989,NULL d803c517c990,NULL d803c517c991,NULL d803c517c992,"{""lat"":null,""lon"":null,""province"":null}" d803c517c993,NULL d803c517c994,NULL d803c517c995,NULL d803c517c996,NULL and i select the result in carbondata use sparksql like this: +------------+---------------------------------------------------+ |device |source | +------------+---------------------------------------------------+ |d803c517c971|[NULL,null,null] | |d803c517c972|[NULL,null,null] | |d803c517c973|[{"lat":null,"lon":null,"province":null},null,null]| |d803c517c974|[NULL,null,null] | |d803c517c975|[NULL,null,null] | |d803c517c976|[NULL,null,null] | |d803c517c977|[NULL,null,null] | |d803c517c978|[{"lat":"22.696963,null,null] | |d803c517c979|[NULL,null,null] | |d803c517c980|[NULL,null,null] | |d803c517c981|[NULL,null,null] | |d803c517c982|[NULL,null,null] | |d803c517c983|[NULL,null,null] | |d803c517c984|[NULL,null,null] | |d803c517c985|[NULL,null,null] | |d803c517c986|[{"lat":"22.542006,null,null] | |d803c517c987|[NULL,null,null] | |d803c517c988|[NULL,null,null] | |d803c517c989|[NULL,null,null] | |d803c517c990|[NULL,null,null] | |d803c517c991|[NULL,null,null] | |d803c517c992|[{"lat":null,"lon":null,"province":null},null,null]| |d803c517c993|[NULL,null,null] | |d803c517c994|[NULL,null,null] | |d803c517c995|[NULL,null,null] | |d803c517c996|[NULL,null,null] | +------------+---------------------------------------------------+ the field source.lon has lost,and data format are not the same as in my csv file. -- This message was sent by Atlassian JIRA (v6.4.14#64029) |
Free forum by Nabble | Edit this page |