Gremlin: Can't sum property from other node - gremlin

I have a graph with post vertices, which have edges to 1 or more tag vertices.
I'm trying to compute the total weight for each tag:
// Create empty graph
graph = TinkerGraph.open()
g = graph.traversal()
// Create some tag vertices
g.inject('food', 'drink').addV('tag').property(id, identity())
// Create some posts with 'food' tag
posts = g.inject(10, 20).addV('post').property('weight', identity()).toList()
g.V(posts).addE('tagged').to(g.V('food'))
// Create some posts with 'drink' tag
posts = g.inject(5).addV('post').property('weight', identity()).toList()
g.V(posts).addE('tagged').to(g.V('drink'))
// Attempt to calculate total weight for each tag
g.V().
hasLabel('post').as('p').
out('tagged').
group('g').
by(id).
by(select('p').values('weight').sum()).
cap('g').next()
but I get this error:
java.lang.Long cannot be cast to org.apache.tinkerpop.gremlin.process.traversal.traverser.util.TraverserSet
Note that computing the total weight for all the posts works fine:
g.V().hasLabel('post').values('weight').sum()
Any suggestions?
Using Gremlin Server 3.3.4

This is a bug in 3.3.4 (and earlier versions). Here's a workaround:
gremlin> g.V().hasLabel('post').as('p').
......1> out('tagged').
......2> group().
......3> by(id).
......4> by(select('p').by('weight').sum())
==>[food:30,drink:5]

Related

Retrieve both edges and vertices in gremlin json

Hi I'm new to gremlin and need to, basically, get everything current to return as a json file. I'm using Syndeia which has a raw query option. I'm having difficulty figuring out how to get both vertices and edges to output. This is what I have thus far:
g.V().has('sLabel','Artifact').has('_isLatest','TRUE').both()
But I also need
g.V().has('sLabel','Artifact').has('_isLatest','TRUE').bothE()
Which gives me all the edges I need, but not the vertices. I realize that g.V() gives vertices and g.E() gives edges, but am unsure how to obtain both/combine them in the one line query. My json comes back as either
"vertices":[] 0 items
"edges":[...] 90 items
Or
"vertices":[...] 90 items
"edges":[] 0 items
My next idea is to do multiple queries into java to parse, combine, and manipulate there. I would really like to avoid doing this, but will if it's the only way. Thank you for being kind and understanding and for your help. I appreciate you all.
You can write a Gremlin query that gives you both the vertices and the edges. There are a few ways to do that. Probably the easiest way is to use path as shown below.
g.V().has('sLabel','Artifact').
has('_isLatest','TRUE').
bothE().
otherV().
path()
as an alternative, you could do:
g.V().has('sLabel','Artifact').
has('_isLatest','TRUE').
bothE().as('e').
otherV().as('v').
select('v','e')
You can do it as follows,
Just see the code below have demonstrated the following,
Vertex query
Edge query
combined query
converted combined query into valueMap
Converted the combined query valueMap into JSON (using version 3)
Converted the combined query valueMap into JSON (using version 1)
Vertex Query
gremlin> g.V(131200).both()
==>v[40992]
==>v[12336]
==>v[20608]
==>v[180376]
gremlin>
Edge query
gremlin> g.V(131200).bothE()
==>e[7rps-2t8g-1fdh-vmo][131200-locatedin->40992]
==>e[3qpy-9io-1bf9-2t8g][12336-containedin->131200]
==>e[7rbk-fwg-1bf9-2t8g][20608-containedin->131200]
==>e[7l03-3v6g-1bf9-2t8g][180376-containedin->131200]
gremlin>
Combined query
gremlin>
gremlin>g.V(131200).project('nodes','edges').by(both().fold()).by(bothE().fold())
==>[nodes:[v[40992],v[12336],v[20608],v[180376]],edges:[e[7rps-2t8g-1fdh-vmo][131200-locatedin->40992],e[3qpy-9io-1bf9-2t8g][12336-containedin->131200],e[7rbk-fwg-1bf9-2t8g][20608-containedin->131200],e[7l03-3v6g-1bf9-2t8g][180376-containedin->131200]]]
gremlin>
converted combined query into valueMap
gremlin>
gremlin> output = g.V(131200).project('nodes','edges').by(both().valueMap().fold()).by(bothE().valueMap().fold()).next()
==>nodes=[{gId=[0b98e8d5-681d-4155-8aaf-5d86babc0cff], isDeleted=[FALSE], isEntity=[TRUE], name=[TotsukaCDC], basetype=[GeoLocation], source=[RCP-Base], type=[DC], geographyL2DisplayName=[KANAGAWA], geographyL2Name=[KANAGAWA], geographyL4Name=[KNG-TOTSUKAKU-NORTH], geographyL4DisplayName=[KNG-TOTSUKAKU-NORTH], displayName=[TotsukaCDC], latitude=[139.525166], locationType=[CDC], geographyL1Name=[KANTO], geographyL3DisplayName=[YOKOHAMA-SHI], geographyL3Name=[YOKOHAMA-SHI], geographyL1DisplayName=[KANTO], locationCode=[tt], longitude=[35.416535]}, {gId=[90e7407c-cd34-44b0-9bf9-1a8866aac429], isDeleted=[FALSE], isEntity=[TRUE], name=[UHN2KNGcdcttRA01_HundredGigE0/0/0/6], basetype=[Entity], source=[RCP-Base], type=[PhysicalPort], portInUse=[TRUE], portNum=[HundredGigE0/0/0/6]}, {gId=[d0fc7284-4cd8-4cba-8d3d-be3c4ca2ba31], isDeleted=[FALSE], isEntity=[TRUE], name=[UHN2KNGcdcttRA01_Hu0/2/0/35], basetype=[Entity], source=[RCP-Base], type=[PhysicalPort], portInUse=[TRUE], portNum=[Hu0/2/0/35]}, {gId=[2f4f5a3a-b88f-4915-acd5-a1d873507f11], isDeleted=[FALSE], isEntity=[TRUE], name=[UHN2KNGcdcttRA01_HundredGigE0/1/0/6], basetype=[Entity], source=[RCP-Base], type=[PhysicalPort], portInUse=[TRUE], portNum=[HundredGigE0/1/0/6]}]
==>edges=[{destName=TotsukaCDC, srcName=UHN2KNGcdcttRA01, gId=1133ff0f-2f4b-4d21-b2d6-fed796ed13f9}, {gId=220d7d2b-8b96-4331-bdc1-50127685946a, destName=UHN2KNGcdcttRA01, srcName=UHN2KNGcdcttRA01_HundredGigE0/0/0/6}, {gId=ebd78fc0-a66c-4033-a776-f80210c02e63, destName=UHN2KNGcdcttRA01, srcName=UHN2KNGcdcttRA01_Hu0/2/0/35}, {srcName=UHN2KNGcdcttRA01_HundredGigE0/1/0/6, gId=0df0f3d0-037a-4caa-9f81-6a6df1c22e25, destName=UHN2KNGcdcttRA01}]
gremlin>
Converted the combined query valueMap into JSON (Using Version 3)
gremlin>
gremlin> mapper = GraphSONMapper.build().version(GraphSONVersion.V3_0).create().createMapper()
==>org.apache.tinkerpop.shaded.jackson.databind.ObjectMapper#10a907ec
gremlin>
gremlin> mapper.writeValueAsString(output) // output is a variable in which data is collected in step 4
==>{"#type":"g:Map","#value":["nodes",{"#type":"g:List","#value":[{"#type":"g:Map","#value":["gId",{"#type":"g:List","#value":["0b98e8d5-681d-4155-8aaf-5d86babc0cff"]},"isDeleted",{"#type":"g:List","#value":["FALSE"]},"isEntity",{"#type":"g:List","#value":["TRUE"]},"name",{"#type":"g:List","#value":["TotsukaCDC"]},"basetype",{"#type":"g:List","#value":["GeoLocation"]},"source",{"#type":"g:List","#value":["RCP-Base"]},"type",{"#type":"g:List","#value":["DC"]},"geographyL2DisplayName",{"#type":"g:List","#value":["KANAGAWA"]},"geographyL2Name",{"#type":"g:List","#value":["KANAGAWA"]},"geographyL4Name",{"#type":"g:List","#value":["KNG-TOTSUKAKU-NORTH"]},"geographyL4DisplayName",{"#type":"g:List","#value":["KNG-TOTSUKAKU-NORTH"]},"displayName",{"#type":"g:List","#value":["TotsukaCDC"]},"latitude",{"#type":"g:List","#value":[{"#type":"g:Double","#value":139.525166}]},"locationType",{"#type":"g:List","#value":["CDC"]},"geographyL1Name",{"#type":"g:List","#value":["KANTO"]},"geographyL3DisplayName",{"#type":"g:List","#value":["YOKOHAMA-SHI"]},"geographyL3Name",{"#type":"g:List","#value":["YOKOHAMA-SHI"]},"geographyL1DisplayName",{"#type":"g:List","#value":["KANTO"]},"locationCode",{"#type":"g:List","#value":["tt"]},"longitude",{"#type":"g:List","#value":[{"#type":"g:Double","#value":35.416535}]}]},{"#type":"g:Map","#value":["gId",{"#type":"g:List","#value":["90e7407c-cd34-44b0-9bf9-1a8866aac429"]},"isDeleted",{"#type":"g:List","#value":["FALSE"]},"isEntity",{"#type":"g:List","#value":["TRUE"]},"name",{"#type":"g:List","#value":["UHN2KNGcdcttRA01_HundredGigE0/0/0/6"]},"basetype",{"#type":"g:List","#value":["Entity"]},"source",{"#type":"g:List","#value":["RCP-Base"]},"type",{"#type":"g:List","#value":["PhysicalPort"]},"portInUse",{"#type":"g:List","#value":["TRUE"]},"portNum",{"#type":"g:List","#value":["HundredGigE0/0/0/6"]}]},{"#type":"g:Map","#value":["gId",{"#type":"g:List","#value":["d0fc7284-4cd8-4cba-8d3d-be3c4ca2ba31"]},"isDeleted",{"#type":"g:List","#value":["FALSE"]},"isEntity",{"#type":"g:List","#value":["TRUE"]},"name",{"#type":"g:List","#value":["UHN2KNGcdcttRA01_Hu0/2/0/35"]},"basetype",{"#type":"g:List","#value":["Entity"]},"source",{"#type":"g:List","#value":["RCP-Base"]},"type",{"#type":"g:List","#value":["PhysicalPort"]},"portInUse",{"#type":"g:List","#value":["TRUE"]},"portNum",{"#type":"g:List","#value":["Hu0/2/0/35"]}]},{"#type":"g:Map","#value":["gId",{"#type":"g:List","#value":["2f4f5a3a-b88f-4915-acd5-a1d873507f11"]},"isDeleted",{"#type":"g:List","#value":["FALSE"]},"isEntity",{"#type":"g:List","#value":["TRUE"]},"name",{"#type":"g:List","#value":["UHN2KNGcdcttRA01_HundredGigE0/1/0/6"]},"basetype",{"#type":"g:List","#value":["Entity"]},"source",{"#type":"g:List","#value":["RCP-Base"]},"type",{"#type":"g:List","#value":["PhysicalPort"]},"portInUse",{"#type":"g:List","#value":["TRUE"]},"portNum",{"#type":"g:List","#value":["HundredGigE0/1/0/6"]}]}]},"edges",{"#type":"g:List","#value":[{"#type":"g:Map","#value":["destName","TotsukaCDC","srcName","UHN2KNGcdcttRA01","gId","1133ff0f-2f4b-4d21-b2d6-fed796ed13f9"]},{"#type":"g:Map","#value":["gId","220d7d2b-8b96-4331-bdc1-50127685946a","destName","UHN2KNGcdcttRA01","srcName","UHN2KNGcdcttRA01_HundredGigE0/0/0/6"]},{"#type":"g:Map","#value":["gId","ebd78fc0-a66c-4033-a776-f80210c02e63","destName","UHN2KNGcdcttRA01","srcName","UHN2KNGcdcttRA01_Hu0/2/0/35"]},{"#type":"g:Map","#value":["srcName","UHN2KNGcdcttRA01_HundredGigE0/1/0/6","gId","0df0f3d0-037a-4caa-9f81-6a6df1c22e25","destName","UHN2KNGcdcttRA01"]}]}]}
gremlin>
Converted the combined query valueMap into JSON (Using Version 1)
gremlin>
gremlin> mapper1 = GraphSONMapper.build().version(GraphSONVersion.V1_0).create().createMapper()
==>org.apache.tinkerpop.shaded.jackson.databind.ObjectMapper#1a7163e3
gremlin>
gremlin> mapper1.writeValueAsString(output)
==>{"nodes":[{"gId":["0b98e8d5-681d-4155-8aaf-5d86babc0cff"],"isDeleted":["FALSE"],"isEntity":["TRUE"],"name":["TotsukaCDC"],"basetype":["GeoLocation"],"source":["RCP-Base"],"type":["DC"],"geographyL2DisplayName":["KANAGAWA"],"geographyL2Name":["KANAGAWA"],"geographyL4Name":["KNG-TOTSUKAKU-NORTH"],"geographyL4DisplayName":["KNG-TOTSUKAKU-NORTH"],"displayName":["TotsukaCDC"],"latitude":[139.525166],"locationType":["CDC"],"geographyL1Name":["KANTO"],"geographyL3DisplayName":["YOKOHAMA-SHI"],"geographyL3Name":["YOKOHAMA-SHI"],"geographyL1DisplayName":["KANTO"],"locationCode":["tt"],"longitude":[35.416535]},{"gId":["90e7407c-cd34-44b0-9bf9-1a8866aac429"],"isDeleted":["FALSE"],"isEntity":["TRUE"],"name":["UHN2KNGcdcttRA01_HundredGigE0/0/0/6"],"basetype":["Entity"],"source":["RCP-Base"],"type":["PhysicalPort"],"portInUse":["TRUE"],"portNum":["HundredGigE0/0/0/6"]},{"gId":["d0fc7284-4cd8-4cba-8d3d-be3c4ca2ba31"],"isDeleted":["FALSE"],"isEntity":["TRUE"],"name":["UHN2KNGcdcttRA01_Hu0/2/0/35"],"basetype":["Entity"],"source":["RCP-Base"],"type":["PhysicalPort"],"portInUse":["TRUE"],"portNum":["Hu0/2/0/35"]},{"gId":["2f4f5a3a-b88f-4915-acd5-a1d873507f11"],"isDeleted":["FALSE"],"isEntity":["TRUE"],"name":["UHN2KNGcdcttRA01_HundredGigE0/1/0/6"],"basetype":["Entity"],"source":["RCP-Base"],"type":["PhysicalPort"],"portInUse":["TRUE"],"portNum":["HundredGigE0/1/0/6"]}],"edges":[{"destName":"TotsukaCDC","srcName":"UHN2KNGcdcttRA01","gId":"1133ff0f-2f4b-4d21-b2d6-fed796ed13f9"},{"gId":"220d7d2b-8b96-4331-bdc1-50127685946a","destName":"UHN2KNGcdcttRA01","srcName":"UHN2KNGcdcttRA01_HundredGigE0/0/0/6"},{"gId":"ebd78fc0-a66c-4033-a776-f80210c02e63","destName":"UHN2KNGcdcttRA01","srcName":"UHN2KNGcdcttRA01_Hu0/2/0/35"},{"srcName":"UHN2KNGcdcttRA01_HundredGigE0/1/0/6","gId":"0df0f3d0-037a-4caa-9f81-6a6df1c22e25","destName":"UHN2KNGcdcttRA01"}]}

Upsert of Vertices and Edge in a single Traversal in Gremlin

I've been trying for several hours to write a gremlin statement to handle upserting 2 vertices and 1 edge between them, without much luck.
in pseudo-gremlin what I want to do is pretty straightforward and is the following:
g.V()
.hasLabel("person")
.has("custom_id", "123")
.fold()
.coalesce(
__.unfold().property(single, "name", "Tom"),
__.addV("person").property(single, "custom_id", "123").property(single, "name", "Tom"))
.as("fromStep")
.V()
.hasLabel("person")
.has("custom_id", "654")
.fold()
.coalesce(
__.unfold().property(single, "name", "Sally"),
__.addV("person").property(single, "custom_id", "654").property(single, "name", "Sally"))
.as("toStep")
.E()
.hasLabel("knows")
.where(__.inV().is("fromStep"))
.where(__.outV().is("toStep"))
.fold()
.coalesce(
__.unfold().property("since", "2020-01-01"),
__.addE("knows").property("since", "2020-01-01").from("fromStep").to("toStep")
The problem with this code is that each fold step, being a barrier step "removes" the value of the previous as steps.
Is there a way to do this properly in a single statement?
SOLUTION
Thanks to Kelvin answer's here's the solution :)
g.V()
.hasLabel("person")
.has("custom_id", "123")
.fold()
.coalesce(
__.unfold().property(single, "name", "Tom"),
__.addV("person").property(single, "custom_id", "123").property(single, "name", "Tom"))
.store("a")
.V()
.hasLabel("person")
.has("custom_id", "654")
.fold()
.coalesce(
__.unfold().property(single, "name", "Sally"),
__.addV("person").property(single, "custom_id", "654").property(single, "name", "Sally"))
.store("b")
.fold()
.select("a").unfold().as("from")
.select("b").unfold().coalesce(
__.inE("knows").where(__.outV().as("from")).property("since", "2020-01-01"),
__.addE("knows").property("since", "2020-01-01").from("from")
)
As you noted, labels applied using an as step are lost after a barrier/map step like fold is used. However, store and aggregate bulk sets will survive intact. Here is a simple pattern that works after a fold that you should be able to adapt for your use case. As we discussed in the comments, if you knew the ID of the vertices you were searching for and if the ID was not found you created it that would give you another way to approach this. Anyway, here is one way that works in the presence of a fold step.
gremlin> g.V('3').store('a').
......1> V('4').store('b').
......2> fold().
......3> select('a').unfold().as('from').
......4> select('b').unfold().as('to').
......5> addE('test').
......6> from('from').
......7> to('to')
==>e[60876][3-test->4]
One other note, in general mid traversal E steps are not supported. It really should be written as V().outE()

what is the best way to search with multi properties in gremlin

I want a path from vertex A,C,D,E,G,H,
I tired with following code
g.V().hasLabel('a').repeat(out().simplePath()).until(hasLabel('h')).
path().by(values('vehicle','time').fold())
my received output
[[10.00,8.30,v1,v2],[10.30,9.00,9.30,v1,v2,v3],[11.00,9.30,10.00,v1,v2,v3],[1
1.30,10.00,10.30,v1,v2,v3],[12.00,11.00,v1,v3],[12.30,11.30,v1,v3]]
expected output
[a:[10.00,v1],c:[11.00,v1],d:[11.30,v1],e:[12.00,v1],g:[12.30,v1],h:[1.00,v1]],[a:[8.30,v2],c:[9.00,v2],d:[9.30,v2],e:[10.00,v2],g:[11.00,v3],h:[11.30,v3]]
the following is a sample graph
v1 = graph.addVertex(id,1,label,'a','vehicle','v1','time',10.00)
v2 = graph.addVertex(id,2,label,'b','vehicle','v1','time',10.30)
v3 = graph.addVertex(id,3,label,'c','vehicle','v1','time',11.00)
v4 = graph.addVertex(id,4,label,'d','vehicle','v1','time',11.30)
v5 = graph.addVertex(id,5,label,'e','vehicle','v1','time',12.00)
v7 = graph.addVertex(id,7,label,'g','vehicle','v1','time',12.30)
v8 = graph.addVertex(id,8,label,'h','vehicle','v1','time',1.00)
v1.addEdge('traveles',v3)
v3.addEdge('traveles',v4)
v4.addEdge('traveles',v5)
v5.addEdge('traveles',v7)
v7.addEdge('traveles',v8)
g.V(1).property(list,'vehicle','v2').property(list,'time',8.30)
g.V(3).property(list,'vehicle','v2').property(list,'time',9.00).property(list,'vehicle','v3').property(list,'time',9.30)
g.V(4).property(list,'vehicle','v2').property(list,'time',9.30).property(list,'vehicle','v3').property(list,'time',10.00)
g.V(5).property(list,'vehicle','v2').property(list,'time',10.00).property(list,'vehicle','v3').property(list,'time',10.30)
g.V(7).property(list,'vehicle','v3').property(list,'time',11.00)
g.V(8).property(list,'vehicle','v3').property(list,'time',11.30)
can anyone help me please
If your property has multiple values then the values step will return them all. Assuming you have two values you can select one or the other using limit or tail.
gremlin> g.addV('test').property(list,'x',1).property(list,'x',2)
==>v[61316]
gremlin> g.V(61316).values('x')
==>1
==>2
gremlin> g.V(61316).values('x').tail(1)
==>2
gremlin> g.V(61316).values('x').limit(1)
==>1
If you have more than 2 values and you need to select one from the middle of a list or set you can use the range step to select that value.
gremlin> g.V(61316).property(list,'x',5)
==>v[61316]
gremlin> g.V(61316).values('x')
==>1
==>2
==>5
gremlin> g.V(61316).values('x').range(1,2)
==>2
All of this said if you could edit your question with some sample data it would help give a better answer as it seems your data model has lots of values for each property so Gremlin is doing what you are asking it to.
Also, in your example you are finding vertices by their label and not testing the values at all. You can test a multi-property just as you would a single property
gremlin> g.V().has('x',2)
==>v[61316
EDITED to add:
Using these techniques, you could change your path statement to something like this:
path().
by(union(label,
values('time').limit(1),
values('vehicle').limit(1)).fold())

How do I collect values from a vertex used in a traversal?

I want the details of a vertex along with details of vertices that are joined to it.
I have a group vertex, incoming 'member' edges to user vertices. I want the details of the vertices.
g.V(1).as('a').in('member').valueMap().as('b').select('a','b').unfold().dedup()
==>a=v[1]
==>b={image=[images/profile/friend9.jpg], name=[Thomas Thompson], email=[me#thomasthompson.co.uk]}
==>b={image=[images/profile/friend13.jpg], name=[Laura Tostevin], email=[me#lauratostevin.co.uk]}
==>b={image=[images/profile/friend5.jpg], name=[Alan Thompson], email=[me#alanthompson.co.uk]}
==>b={image=[images/profile/friend10.jpg], name=[Laura Bourne], email=[me#laurabourne.co.uk]}
Ideally what I'd want is:
{label: 'group', id=1, name='A Group', users=[{id=2, label="user",name=".."}, ... }]}
When I tried a project, it didn't like me using 'in'
gremlin> g.V('1').project('name','users').by('name').by(in('member').select())
groovysh_parse: 1: unexpected token: in # line 1, column 83.
'name','users').by('name').by(in('member
To get your preferred output format, you have to join the group's valueMap() with the list of users. On TinkerPop's modern toy graph you would do something like this:
gremlin> g.V(3).union(valueMap(true).
by(unfold()),
project('users').
by(__.in('created').
valueMap(true).
by(unfold()).
fold())).
unfold().
group().
by(keys).
by(select(values))
==>[name:lop,id:3,lang:java,label:software,users:[[id:1,label:person,name:marko,...],...]]
Mapping this to your graph should be pretty straight-forward, it's basically just about changing labels.
Because in is a reserved keyword in Groovy you must use the verbose syntax __.in
try:
g.V('1').project('name','users').by('name').by(__.in('member').valueMap(true).fold())

Gremlin: remove repeated parent edges

My Query:
g.V().has("a","123").has("s","1").has("e","p").has("r","e-1").repeat(outE().where(values("startTime").is(gte("2018-12-15"))).where(values("endTime").is(lte("2018-12-16"))).otherV().simplePath()).emit().path().by(valueMap("a","s")).fold()
I am getting result as
A-->B
A-->B-->C
A-->B-->C-->D
How do I remove the first 2 lines from the output?
Desired output: A-->B-->C-->D
Thanks
Don't emit the paths in the first place.
edgeTraversal = outE().
has("startTime", gte("2018-12-15")).
has("endTime", lte("2018-12-16")).simplePath(); []
g.V().has("a","123").has("s","1").has("e","p").has("r","e-1").
repeat(edgeTraversal.clone().inV()).
until(__.not(edgeTraversal)).
path().
by(valueMap("a","s")).fold()
Without the child traversal variable:
g.V().has("a","123").has("s","1").has("e","p").has("r","e-1").
repeat(outE().
has("startTime", gte("2018-12-15")).
has("endTime", lte("2018-12-16")).simplePath().inV()).
until(__.not(outE().
has("startTime", gte("2018-12-15")).
has("endTime", lte("2018-12-16")).simplePath())).
path().
by(valueMap("a","s")).fold()

Resources