Hi Val,
In order to run the HepPlanner you only need to execute
HepPlanner#findBestExp() to obtain the transformed plan (no need to pass it
also to a program which will call the same method again). Note also that
planner instances are meant to be used only once since they are stateful,
so you shou
None of your 4 points prove that anything is fundamentally broken.
In particular, the ability to represent cyclic graphs is a feature, not a bug.
It’s not “too late” if a rule fires and creates a RelNode and that RelNode’s
constructor asserts. Because that rule is being run in a debugger, as par
Siddharth Teotia created CALCITE-2831:
-
Summary: Fix EXTRACT(SECOND FROM )
Key: CALCITE-2831
URL: https://issues.apache.org/jira/browse/CALCITE-2831
Project: Calcite
Issue Type: Bug
Siddharth Teotia created CALCITE-2830:
-
Summary: Fix AvgVarianceConvertlet
Key: CALCITE-2830
URL: https://issues.apache.org/jira/browse/CALCITE-2830
Project: Calcite
Issue Type: Bug
Siddharth Teotia created CALCITE-2829:
-
Summary: Fix RexSimplify#processRange cast issue
Key: CALCITE-2829
URL: https://issues.apache.org/jira/browse/CALCITE-2829
Project: Calcite
Issue T
Siddharth Teotia created CALCITE-2828:
-
Summary: Handle cost propagation properly in Volcano Planner
Key: CALCITE-2828
URL: https://issues.apache.org/jira/browse/CALCITE-2828
Project: Calcite
Siddharth Teotia created CALCITE-2827:
-
Summary: Allow CONVENTION.NONE planning with VolcanoPlanner
Key: CALCITE-2827
URL: https://issues.apache.org/jira/browse/CALCITE-2827
Project: Calcite
Siddharth Teotia created CALCITE-2823:
-
Summary: Prevent optimization of non-deterministic functions
Key: CALCITE-2823
URL: https://issues.apache.org/jira/browse/CALCITE-2823
Project: Calcite
Siddharth Teotia created CALCITE-2826:
-
Summary: Add versions of create() for LogicalSort and
LogicalExchange that take cluster and trait set as input
Key: CALCITE-2826
URL: https://issues.apache.org/jira/brow
Siddharth Teotia created CALCITE-2825:
-
Summary: Add function double power (BigDecimal b0, long b1)
Key: CALCITE-2825
URL: https://issues.apache.org/jira/browse/CALCITE-2825
Project: Calcite
Siddharth Teotia created CALCITE-2824:
-
Summary: Fix invalid usage of RexExecutorImpl
Key: CALCITE-2824
URL: https://issues.apache.org/jira/browse/CALCITE-2824
Project: Calcite
Issue Type
Siddharth Teotia created CALCITE-2822:
-
Summary: Allow multi-join rules to be used with any project/filter
Key: CALCITE-2822
URL: https://issues.apache.org/jira/browse/CALCITE-2822
Project: Calcite
Siddharth Teotia created CALCITE-2821:
-
Summary: Improve extensibility of Volcano Planner
Key: CALCITE-2821
URL: https://issues.apache.org/jira/browse/CALCITE-2821
Project: Calcite
Issue
Siddharth Teotia created CALCITE-2819:
-
Summary: Add a version of LoptOptimizeJoinRule that uses the first
ordering it finds
Key: CALCITE-2819
URL: https://issues.apache.org/jira/browse/CALCITE-2819
Siddharth Teotia created CALCITE-2820:
-
Summary: Add a version of AggregateReduceFunctionsRule that does
not reduce SUm -> SUM0
Key: CALCITE-2820
URL: https://issues.apache.org/jira/browse/CALCITE-2820
This looks like a bug. Could you try adding another key "userConfig":
"{}" under "operand" in your model file and see if that runs?
--
Michael Mior
mm...@apache.org
Le lun. 4 févr. 2019 à 14:29, Allan Keers a écrit :
>
> Hi there,
> I can connect my Calcite installation to the sample csv and que
Hi there,
I can connect my Calcite installation to the sample csv and query the tables.
When I try to connect to Elasticsearch I get the following error.
Any help would be appreciated.
Thanks,
Allan Keers
sqlline version 1.6.0
sqlline> !connect
jdbc:calcite:model=/Users/Administrator/dev/calcit
RexShuttle does not have a watertight specification. It does what it needs to
do for the cases that use it. Generally, if the sub-elements of a RelNode are
sub-types of RexNode, RexShuttle processes them, otherwise it ignores them. But
pragmatically, as long as what you do doesn’t break the test
Michael is correct that to get all alternative plans you can create your own
visitor.
However note that the number of plans is typically very large. Consider a query
that starts of as
Scan -> Project -> Filter
and broadens to 3 equivalence sets with 3 RelNodes in each:
{R11, R12, R13} ->
I'm not convinced there aren't better alternatives, so I hope we can
continue to explore this. Specifically with respect to the statement
"It means there's not enough rules to produce the node with desired
properties", it would be great if we could more clearly identify what
properties the missing
Great idea. I was intending to volunteer as RM last time, but with the
time pressure, I didn't respond soon enough. I'm happy to take the
April release (1.20).
--
Michael Mior
mm...@apache.org
Le jeu. 31 janv. 2019 à 18:54, Andrei Sereda a écrit :
>
> Release Target date Release manager
> ==
In general, there's no way to guarantee you get "all" the plans for a
particular query. If you want to get a variety of plans and their
associated cost, take a look at CheapestPlanVisitor in RelSubset.java.
This is used by VolcanoPlanner to build the cheapest plan at the end
of the planning phase.
Mickaël Sauvée created CALCITE-2818:
---
Summary: Implementation of Extract (ExtractImplementor) returns
wrong result for dates before Epoch.
Key: CALCITE-2818
URL: https://issues.apache.org/jira/browse/CALCITE-28
The Apache Jenkins build system has built Calcite-Master (build #1007)
Status: Still Failing
Check console output at https://builds.apache.org/job/Calcite-Master/1007/ to
view the results.
Hi Stamatis,
Thank you for the answer! I have some questions to clarify that.
1. What if my new node doesn’t strictly have a `RexNode`, but it’s referring to
the inputs indirectly, like via `String fieldName` instead of using RexInputRef?
2. Arguably this is the similar thing what `Aggregate` no
25 matches
Mail list logo
