[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[ASF GitHub Bot (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16389680#comment-16389680
 ] 

ASF GitHub Bot commented on TINKERPOP-1777:
---

Github user asfgit closed the pull request at:

https://github.com/apache/tinkerpop/pull/805


> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>  Components: process
>Affects Versions: 3.2.6
>Reporter: Sebastian Estevez
>Assignee: Daniel Kuppitz
>Priority: Major
>  Labels: breaking
>
> To reproduce:
> {code}gremlin> g.V().values('test').max()
> ==>-2147483648{code}
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[ASF GitHub Bot (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16389679#comment-16389679
 ] 

ASF GitHub Bot commented on TINKERPOP-1777:
---

Github user dkuppitz commented on a diff in the pull request:

https://github.com/apache/tinkerpop/pull/805#discussion_r172874794
  
--- Diff: gremlin-test/features/branch/Union.feature ---
@@ -135,7 +135,6 @@ Feature: Step - union()
   | d[3].l   |
   | d[0].l   |
   | d[1.9].d |
-  | d[0].i   |
--- End diff --

> just so i'm clear, this value is no longer in the output because vadas 
(aka v[2]) has no out edges to sum a weight on...is that right?

Exactly.

It's the same thing as doing just `g.V(2).outE()` - the traverser would 
simply die as there are no outgoing edges. Likewise with 
`outE().values('weight').sum()`, there simply is no sum, hence the traverser 
dies.


> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>  Components: process
>Affects Versions: 3.2.6
>Reporter: Sebastian Estevez
>Assignee: Daniel Kuppitz
>Priority: Major
>  Labels: breaking
>
> To reproduce:
> {code}gremlin> g.V().values('test').max()
> ==>-2147483648{code}
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[ASF GitHub Bot (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16388358#comment-16388358
 ] 

ASF GitHub Bot commented on TINKERPOP-1777:
---

Github user spmallette commented on a diff in the pull request:

https://github.com/apache/tinkerpop/pull/805#discussion_r172629315
  
--- Diff: gremlin-test/features/branch/Union.feature ---
@@ -135,7 +135,6 @@ Feature: Step - union()
   | d[3].l   |
   | d[0].l   |
   | d[1.9].d |
-  | d[0].i   |
--- End diff --

just so i'm clear, this value is no longer in the output because vadas (aka 
`v[2]`) has no out edges to sum a weight on...is that right? if so, do you 
happen to know what catches the thrown `FastNoSuchElementException` and 
prevents it from bubbling up just kill the whole traversal? (i'm just trying to 
follow how the code works here for my own edification)


> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>  Components: process
>Affects Versions: 3.2.6
>Reporter: Sebastian Estevez
>Assignee: Daniel Kuppitz
>Priority: Major
>  Labels: breaking
>
> To reproduce:
> {code}gremlin> g.V().values('test').max()
> ==>-2147483648{code}
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[ASF GitHub Bot (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16388341#comment-16388341
 ] 

ASF GitHub Bot commented on TINKERPOP-1777:
---

Github user spmallette commented on a diff in the pull request:

https://github.com/apache/tinkerpop/pull/805#discussion_r172627151
  
--- Diff: docs/src/upgrade/release-3.4.x.asciidoc ---
@@ -28,3 +28,45 @@ NEED AN IMAGE
 Please see the 
link:https://github.com/apache/tinkerpop/blob/3.4.0/CHANGELOG.asciidoc#release-3-4-0[changelog]
 for a complete list of all the modifications that are part of this release.
 
 === Upgrading for Users
+
+ Modifications to reducing barrier steps
+
+The behavior of `min()`, `max()`, `mean()` and `sum()` has been modified 
to return no result if there's no input. Previously these steps yielded the 
internal seed value:
+
+[source,groovy]
+
+gremlin> g.V().values('foo').min()
+==>NaN
+gremlin> g.V().values('foo').max()
+==>NaN
+gremlin> g.V().values('foo').mean()
+==>NaN
+gremlin> g.V().values('foo').sum()
+==>0
+
+
+These traversals will no longer emit a result. Note, that this also 
affects more complex scenarios, e.g. if these steps are used in `by()` 
modulators:
+
+[source,groovy]
+
+gremlin> g.V().group().
+..1>   by(label).
+..2>   by(outE().values("weight").sum())
+==>[software:0,person:3.5]
+
+
+Since software vertices have no outgoing edges and thus no weight values 
to sum, `software` will no longer show up in the result. In order to get the 
same result as before, one would
+have to add a `coalesce()`-step:
+
+[source,groovy]
+
+gremlin> g.V().group().
+..1>   by(label).
+..2>   by(outE().values("weight").sum())
+==>[person:3.5]
+gremlin> 
+gremlin> g.V().group().
+..1>   by(label).
+..2>   by(coalesce(outE().values("weight"), constant(0)).sum())
+==>[software:0,person:3.5]
+
--- End diff --

please add the "See: {jira_issue}"


> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>  Components: process
>Affects Versions: 3.2.6
>Reporter: Sebastian Estevez
>Assignee: Daniel Kuppitz
>Priority: Major
>  Labels: breaking
>
> To reproduce:
> {code}gremlin> g.V().values('test').max()
> ==>-2147483648{code}
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[ASF GitHub Bot (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16384423#comment-16384423
 ] 

ASF GitHub Bot commented on TINKERPOP-1777:
---

Github user spmallette commented on a diff in the pull request:

https://github.com/apache/tinkerpop/pull/805#discussion_r172001752
  
--- Diff: CHANGELOG.asciidoc ---
@@ -25,6 +25,8 @@ NEED AND IMAGE
 
 This release also includes changes from <>.
 
+* Fixed a bug in `ReducingBarrierStep`, that returned the provided seed 
value despite no elements being available.
--- End diff --

Should this be a changelog entry for 3.4.0 as this is pointed at the master 
branch?


> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>  Components: process
>Affects Versions: 3.2.6
>Reporter: Sebastian Estevez
>Assignee: Daniel Kuppitz
>Priority: Major
>  Labels: breaking
>
> To reproduce:
> {code}gremlin> g.V().values('test').max()
> ==>-2147483648{code}
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[ASF GitHub Bot (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16384158#comment-16384158
 ] 

ASF GitHub Bot commented on TINKERPOP-1777:
---

GitHub user dkuppitz opened a pull request:

https://github.com/apache/tinkerpop/pull/805

TINKERPOP-1777 Gremlin .max step returns -2147483648 for empty result sets

https://issues.apache.org/jira/browse/TINKERPOP-1777

Fixed the behavior of `min()`, `max()`, `mean()` and `sum()`.
If no input is given, these steps will now throw a `FastNoElementException`.

Also adjusted existing test cases to be in line with the new behavior.

`docker/build.sh -t -i` passed.

VOTE: +1

You can merge this pull request into a Git repository by running:

$ git pull https://github.com/apache/tinkerpop TINKERPOP-1777

Alternatively you can review and apply these changes as the patch at:

https://github.com/apache/tinkerpop/pull/805.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

This closes #805


commit 32014426f12478e13d73b3c1bdca8df6d79d08da
Author: Daniel Kuppitz 
Date:   2018-02-27T17:15:07Z

Fixed the behavior of `min()`, `max()`, `mean()` and `sum()`.
If no input is given, these steps will now throw a `FastNoElementException`.

Also adjusted existing test cases to be in line with the new behavior.

commit 7149f8e8c1c255420b19748632ee4d928480f2e2
Author: Daniel Kuppitz 
Date:   2018-03-02T21:04:58Z

Updated release docs




> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>  Components: process
>Affects Versions: 3.2.6
>Reporter: Sebastian Estevez
>Assignee: Daniel Kuppitz
>Priority: Major
>  Labels: breaking
>
> To reproduce:
> {code}gremlin> g.V().values('test').max()
> ==>-2147483648{code}
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[Daniel Kuppitz (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16350700#comment-16350700
 ] 

Daniel Kuppitz commented on TINKERPOP-1777:
---

{{FastNoSuchElementException}} should actually be expected. There's a bug in 
{{ReducingBarrierStep}} that assigns the seed value although there are no 
elements. Easy one line fix, but the other {{NaN}} issue I mentioned is 
trickier than I thought.

> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>  Components: process
>Affects Versions: 3.2.6
>Reporter: Sebastian Estevez
>Priority: Major
>
> To reproduce:
> {code}gremlin> g.V().values('test').max()
> ==>-2147483648{code}
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[Daniel Kuppitz (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16350589#comment-16350589
 ] 

Daniel Kuppitz commented on TINKERPOP-1777:
---

Hmm, with the latest changes it returns {{NaN}} for empty traversers and does 
throw an exception for non-numeric types.

{noformat}
gremlin> g.V().values('age').max()
==>35
gremlin> g.V().values('weight').max()
==>NaN
gremlin> g.V().values('name').max()
java.lang.String cannot be cast to java.lang.Number
{noformat}

I wouldn't want to make any more changes  in {{NumberHelper}}, but I'd be good 
with the respective steps throwing {{FastNoSuchElementExceptions}}, in case 
{{NumberHelper}} returns {{NaN}}.

Now, after doing a few tests, I'm also wondering why this is happening:

{noformat}
gremlin> __.inject(Double.NaN).max()
==>0
{noformat}

> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>  Components: process
>Affects Versions: 3.2.6
>Reporter: Sebastian Estevez
>Priority: Major
>
> To reproduce:
> {code}gremlin> g.V().values('test').max()
> ==>-2147483648{code}
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (TINKERPOP-1777) Gremlin .max step returns -2147483648 for empty result sets

[Marko A. Rodriguez (JIRA)]

[ 
https://issues.apache.org/jira/browse/TINKERPOP-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16161914#comment-16161914
 ] 

Marko A. Rodriguez commented on TINKERPOP-1777:
---

Perhaps throw a {{FastNoSuchElementException}}.

> Gremlin .max step returns -2147483648 for empty result sets
> ---
>
> Key: TINKERPOP-1777
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1777
> Project: TinkerPop
>  Issue Type: Bug
>Reporter: Sebastian Estevez
>
> To reproduce:
> 
> gremlin> g.V().values('test').max()
> ==>-2147483648
> 
> This should probably return an exeption, "cannot take max of nothing"



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)