# Multiple Values for a property

**URL:** <https://community.neo4j.com/t/multiple-values-for-a-property/52766>\
**Category:** Neo4j Graph Platform\
**Created:** [March 2, 2022, 8:27am UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766 "2022-03-02T08:27:35Z")\
**Posts on this page:** 9\
**Page:** 1

<div class="post-metadata">

**Author:** ![ask.singh90](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/ask.singh90/32/21188_2.png) [@ask.singh90](https://community.neo4j.com/u/ask.singh90)\
**Post date:** [March 2, 2022, 8:27am UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/1 "2022-03-02T08:27:35Z")

</div>

Hi, I am a newbie to Neo4j and struck with following problem.  
My Dataset is as below  
Object, Size, Color  
A,200,White  
A,300,Black  
A,200,Pink  
B,300,White  
B,300,Black

The expected output is two nodes A and B with two properties Size and Color but there should not be duplicate size being displayed. For Example Node A has 200 in White and Pink Color but Size property should be showing 200 only once i.e. (200,300) and Color property should be showing (White, Black, Pink). Can somebody please help me with cypher code?

---

<div class="post-metadata">

**Author:** ![bennu\_neo](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/bennu_neo/32/27616_2.png) [@bennu\_neo](https://community.neo4j.com/u/bennu_neo)\
**Post date:** [March 2, 2022, 10:02am UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/2 "2022-03-02T10:02:55Z")

</div>

Hi @ask.singh90 ,

How have you modelled this problem? Or is it actually that the question?

Are you planning on creating instances of each object with property/tag the gives you type while you have specific size color combinations on each one of them? Dense Object nodes with _HAS\_SIZE_ and _HAS\_COLOR_ to other dense nodes?

Bennu

---

<div class="post-metadata">

**Author:** ![ask.singh90](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/ask.singh90/32/21188_2.png) [@ask.singh90](https://community.neo4j.com/u/ask.singh90)\
**Post date:** [March 2, 2022, 10:30am UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/3 "2022-03-02T10:30:53Z")

</div>

I want to show just the object nodes with multiple values for a property. like A with Size property(200,300) and Color property(White, Black, Pink).

---

<div class="post-metadata">

**Author:** ![bennu\_neo](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/bennu_neo/32/27616_2.png) [@bennu\_neo](https://community.neo4j.com/u/bennu_neo)\
**Post date:** [March 2, 2022, 10:38am UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/4 "2022-03-02T10:38:27Z")

</div>

Hi @ask.singh90 ,

Based on a data like :

```auto
CREATE (o:OBJECT {type : 'A', size : 200, color : 'White'});
CREATE (o:OBJECT {type : 'A', size : 300, color : 'Black'});
CREATE (o:OBJECT {type : 'A', size : 200, color : 'Pink'});
CREATE (o:OBJECT {type : 'B', size : 300, color : 'White'});
CREATE (o:OBJECT {type : 'B', size : 300, color : 'Black'});

```

You can get the expected result with:

```auto
MATCH(o:OBJECT)
WITH o.type as type, collect(distinct o.size) as sizes, collect(distinct o.color) as colors
RETURN type, sizes, colors

```

Notice that it may be misleading because you don't have the object A on size 300 color pink.

Bennu

---

<div class="post-metadata">

**Author:** ![ask.singh90](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/ask.singh90/32/21188_2.png) [@ask.singh90](https://community.neo4j.com/u/ask.singh90)\
**Post date:** [March 2, 2022, 11:02am UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/5 "2022-03-02T11:02:39Z")

</div>

Hi bennu - thanks for your reply but I want to make the generic query. I have a spreadsheet with thousands of rows like this. I wrote following code but the problem with it is, it keeps on adding new value irrespective of whether it is already present or not so A gets size property(200,300,200) instead of expected (200,300)  
LOAD CSV WITH HEADERS FROM "file:///myfile.csv" AS nodeRecord  
MERGE (n: `Object` { `object`:nodeRecord.object })  
on CREATE SET n.`size` = [nodeRecord.`size`]  
on CREATE SET n.`color` = [nodeRecord.`color`]  
on MATCH SET n.`size` = n.`size` + [nodeRecord.`size`]  
on MATCH SET n.`color` = n.`color` + [nodeRecord.`color`]

---

<div class="post-metadata">

**Author:** ![bennu\_neo](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/bennu_neo/32/27616_2.png) [@bennu\_neo](https://community.neo4j.com/u/bennu_neo)\
**Post date:** [March 2, 2022, 1:20pm UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/6 "2022-03-02T13:20:13Z")

</div>

Hi @ask.singh90 ,

Are planning on use APOC on your application?

This may help you

> **[apoc.coll.toSet - APOC Extended Documentation](https://neo4j.com/labs/apoc/4.1/overview/apoc.coll/apoc.coll.toSet/)**
>
> This section contains reference documentation for the apoc.coll.toSet function.

Bennu

---

<div class="post-metadata">

**Author:** ![ask.singh90](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/ask.singh90/32/21188_2.png) [@ask.singh90](https://community.neo4j.com/u/ask.singh90)\
**Post date:** [March 2, 2022, 1:27pm UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/7 "2022-03-02T13:27:00Z")

</div>

Hi Bennu, I am sorry I don't know much about usage of apoc. I am just doing a small poc to learn more about neo4j and cypher and my data set has values as i mentioned in my original post.

---

<div class="post-metadata">

**Author:** ![glilienfield](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/glilienfield/32/27534_2.png) [@glilienfield](https://community.neo4j.com/u/glilienfield)\
**Post date:** [March 2, 2022, 4:17pm UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/8 "2022-03-02T16:17:04Z")

</div>

I don't see anything wrong with your query that would cause duplicate A and B nodes. I run it locally and got a single A node and a single B node, each with the aggregated data.

As a note, the query creates size and color lists that contain duplicate values because you are pushing new values to each list.

You could use a query like the following to avoid this. The query processes the entire import file first to calculate the aggregates of distinct values, then creates the nodes. Just a note, more memory would be required if you have huge files.

LOAD CSV WITH HEADERS FROM "file:///myfile.csv" AS nodeRecord  
with nodeRecord.object as object, collect(distinct nodeRecord.size) as size, collect(distinct nodeRecord.color) as color  
merge (n: Object { object:object }) set n.size = size, n.color = color

---

<div class="post-metadata">

**Author:** ![ask.singh90](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/ask.singh90/32/21188_2.png) [@ask.singh90](https://community.neo4j.com/u/ask.singh90)\
**Post date:** [March 2, 2022, 4:23pm UTC](https://community.neo4j.com/t/multiple-values-for-a-property/52766/9 "2022-03-02T16:23:51Z")

</div>

Thanks @glilienfield ..it works well