as(), where() and some other steps need a place to store a state of the treverser.
as(), where() and some other steps need a place to store a state of the treverser. So we add "~traversers" cstruct column to hold reference to it. This method is used by as() step to register nemed traverser
name provided to as(label) call
Dataframe with current state
Dataframe with label column added to the ~traversers that point ot the current df.
add steps from another traversal, the first graph step is skipped.
the main processing method.
the main processing method. It calls interceptors then does default processing
traversal result as a DataFrame
val find result step by skipping sorting and dedup
val find result step by skipping sorting and dedup
Use this method to get Step properties to work around interface inconsistency.
Use this method to get Step properties to work around interface inconsistency.
Set name of the graph in case it was lost during conversion operations or the result should be stored in another graph The name will be updated in the DseGraphFrame
Set name of the graph in case it was lost during conversion operations or the result should be stored in another graph The name will be updated in the DseGraphFrame
of the graph
this
Iterator methods cause DataFrame collect() call, that materializes all rows on the driver call at the end of computations.
Iterator methods cause DataFrame collect() call, that materializes all rows on the driver call at the end of computations.
Check the step is required only id column The method is used in traversal steps to reduce number of join operations
Check the step is required only id column The method is used in traversal steps to reduce number of join operations
to check
true for steps that need only vertex id
Find OrStep and AndStep in infix notation and rearrange them to prefix one so leftSteps().and().rightSteps() => and(leftSteps(), rightSteps())
Find OrStep and AndStep in infix notation and rearrange them to prefix one so leftSteps().and().rightSteps() => and(leftSteps(), rightSteps())
rearranged steps
Iterator methods cause DataFrame collect() call, that materializes all rows on the driver only count().next() is supported for now It is recommended to get table like results with df() method.
Iterator methods cause DataFrame collect() call, that materializes all rows on the driver only count().next() is supported for now It is recommended to get table like results with df() method.
check that steps inside LocalStep are safe to ignore local step, (has no order, dedupe, limit, etc steps) It uses white list of supported safe stpes.
check that steps inside LocalStep are safe to ignore local step, (has no order, dedupe, limit, etc steps) It uses white list of supported safe stpes.
to check
Iterator methods cause DataFrame collect() call, that materializes all rows on the driver call at the end of computations.
Iterator methods cause DataFrame collect() call, that materializes all rows on the driver call at the end of computations.
there are two combination of vertex steps Vertex->Vertex | g.V().out() Vertex->Egde | g.V().outE() with three direcctions: IN, OUT, BOTH
there are two combination of vertex steps Vertex->Vertex | g.V().out() Vertex->Egde | g.V().outE() with three direcctions: IN, OUT, BOTH
there are three combination of edge steps g.E().outV() g.E().inV() g.E().bothV() with three directions: IN, OUT, BOTH
there are three combination of edge steps g.E().outV() g.E().inV() g.E().bothV() with three directions: IN, OUT, BOTH
if standard id is used and user do not use it as regular properties in some vertices.
if standard id is used and user do not use it as regular properties in some vertices. valueMap and values() should not show them.
"community_id", "member_id" or both if they are not used as regular properties
traversal is mutable when steps are applied the typeTag is used to control step to step safety and type of the last step result for the iterator.
wrap vertex data frame to implement TinkerPop functions