== Physical Plan == CollectLimit (7) +- * ColumnarToRow (6) +- InMemoryTableScan (1) +- InMemoryRelation (2) +- * Project (5) +- * Filter (4) +- Scan csv (3) (1) InMemoryTableScan Output [2]: [turnover#94271745, days_hold#94271808] Arguments: [turnover#94271745, days_hold#94271808] (2) InMemoryRelation Arguments: [turnover#94271745, days_hold#94271808], CachedRDDBuilder(org.apache.spark.sql.execution.columnar.DefaultCachedBatchSerializer@208e3fd9,StorageLevel(disk, memory, deserialized, 1 replicas),*(1) Project [CASE WHEN ((turnover#94271563 = NA) OR (turnover#94271563 = null)) THEN null ELSE cast(turnover#94271563 as float) END AS turnover#94271745, CASE WHEN ((turnover#94271563 = NA) OR (turnover#94271563 = null)) THEN null ELSE (1.0 / cast(cast(turnover#94271563 as float) as double)) END AS days_hold#94271808] +- *(1) Filter ((isnotnull(cap#94271532) AND NOT coalesce(((cap#94271532 = NA) OR (cap#94271532 = null)), false)) AND (cast(cap#94271532 as float) = 0.0)) +- FileScan csv [cap#94271532,turnover#94271563] Batched: false, DataFilters: [isnotnull(cap#94271532), NOT coalesce(((cap#94271532 = NA) OR (cap#94271532 = null)), false), (c..., Format: CSV, Location: InMemoryFileIndex(1 paths)[file:/srv/plusamp/data/default/ea-market/output/digital_revenue_signal..., PartitionFilters: [], PushedFilters: [IsNotNull(cap)], ReadSchema: struct<cap:string,turnover:string> ,None) (3) Scan csv Output [2]: [cap#94271532, turnover#94271563] Batched: false Location: InMemoryFileIndex [file:/srv/plusamp/data/default/ea-market/output/digital_revenue_signal/rev_signal_percentile_r_100/stats_overall.csv] PushedFilters: [IsNotNull(cap)] ReadSchema: struct<cap:string,turnover:string> (4) Filter [codegen id : 1] Input [2]: [cap#94271532, turnover#94271563] Condition : ((isnotnull(cap#94271532) AND NOT coalesce(((cap#94271532 = NA) OR (cap#94271532 = null)), false)) AND (cast(cap#94271532 as float) = 0.0)) (5) Project [codegen id : 1] Output [2]: [CASE WHEN ((turnover#94271563 = NA) OR (turnover#94271563 = null)) THEN null ELSE cast(turnover#94271563 as float) END AS turnover#94271745, CASE WHEN ((turnover#94271563 = NA) OR (turnover#94271563 = null)) THEN null ELSE (1.0 / cast(cast(turnover#94271563 as float) as double)) END AS days_hold#94271808] Input [2]: [cap#94271532, turnover#94271563] (6) ColumnarToRow [codegen id : 1] Input [2]: [turnover#94271745, days_hold#94271808] (7) CollectLimit Input [2]: [turnover#94271745, days_hold#94271808] Arguments: 1000000