So let's say you are able to plot out an object through time using a dataflow loop. You would have a set of data (eg. a snapshot of open opportunities every day that runs every5am.) In addition to plotting this time-series, you want to plot changes from snapshot Date1 to snapshot Date 2 grouped by a certain field like per zip code, or region, or business unit,etc.
 Below is a snip of the waterfall chart and its SAQL
q = load \"zds_IoTForecast5\";\nstartVal = filter q by {{column(qDateTest_1.selection,[\"cTimeStampText\"]).asEquality('cTimeStampText')}};\nendVal = filter q by {{column(q_SnapDate2_1.selection,[\"cTimeStampText\"]).asEquality('cTimeStampText')}};\nq_A = filter q by {{column(qDateTest_1.selection,[\"cTimeStampText\"]).asEquality('cTimeStampText')}};\nq_B = filter q by {{column(q_SnapDate2_1.selection,[\"cTimeStampText\"]).asEquality('cTimeStampText')}};\nstartVal = group startVal by all;\nstartVal = foreach startVal generate \"Start Value\" as '{{column(static_6.selection,[\"Valu\"]).asObject()}}',number_to_string(sum (startVal.'TotValueUSD'),\"$#,###\") as 'Forecast Delta';\nendVal = group endVal by all;\nresult = group q_A by '{{column(static_6.selection,[\"Valu\"]).asObject()}}' full, q_B by '{{column(static_6.selection,[\"Valu\"]).asObject()}}';\nresult = foreach result generate coalesce(q_A.'{{column(static_6.selection,[\"Valu\"]).asObject()}}', q_B.'{{column(static_6.selection,[\"Valu\"]).asObject()}}') as '{{column(static_6.selection,[\"Valu\"]).asObject()}}',coalesce(sum(q_A.'TotValueUSD'),0) as 'A', coalesce(sum(q_B.'TotValueUSD'),0) as 'B';\nresult = foreach result generate '{{column(static_6.selection,[\"Valu\"]).asObject()}}',number_to_string( B-A,\"$#,###\") as 'Forecast Delta';\nendVal = foreach endVal generate \"Ending Value\" as '{{column(static_6.selection,[\"Valu\"]).asObject()}}',number_to_string(sum (endVal.'TotValueUSD'),\"$#,###\") as 'Forecast Delta';\nresult = order result by '{{column(static_6.selection,[\"Valu\"]).asObject()}}' asc;\nfinal = union startVal,result,endVal;