The "pattern" for MERGE in this case is just label + id, so that is exactly
what you want.
you would use this one then:
> MERGE (n:Label {id:{id}})
> SET n.foo = {foo}, n.bar = {bar}
or for row based
> LOAD CSV WITH HEADERS FROM "URL" as row
> MERGE (n:Label {id:row.id <http://row.id/>})
> SET n += row
There would be an option where you first check for the fields that have changed
and only updates those, but I'm not sure it's worth it.
Something like
LOAD CSV WITH HEADERS FROM "URL" as row
MERGE (n:Label {id:row.id <http://row.id/>})
FOREACH (key IN filter(key IN keys(row) WHERE row[key] <> n[key]) | SET n[key]
= row[key] )
Michael
This is exactly what I meant with the
> Am 27.01.2016 um 12:56 schrieb Pranab Banerjee <[email protected]>:
>
> Thank you Michael
>
> I am new to Neo4j and as per my knowledge "MERGE" will not partially use
> existing patterns- its ALL or Nothing. that means either whole patterns
> matches or whole patter created.
>
> for example like below, I am not sure whether MERGE will be effective ??
>
>
> One time load (/user/home/onetime.csv)-
> =============================================
> Id, field1, field2, field3, last_modified_dt
> === ====== ====== ======= =================
> 1 abcd efgh 12 2016-01-25 09:22:03 <-(New entry)
> 2 mnop efgh 14 2016-01-25 09:22:04 <-(New entry)
>
> After loading the above onetime data into respective node we received the
> below incremental data
>
> Incremental Load (/user/home/incremental.csv)-
> ===============================================
> Id, field1, field2, field3, last_modified_dt
> === ====== ====== ======= =================
> 2 txyz efgh 18 2016-01-27 09:48:03 <-(modified data)
> 3 hijk octu 17 2016-01-27 09:49:00 <-(New entry)
>
> Appreciate any suggestion on this. Thanks in advance.
>
> Pranab
>
>
>
> On Wed, Jan 27, 2016 at 4:50 PM, Michael Hunger
> <[email protected] <mailto:[email protected]>>
> wrote:
> If you have a timestamp or other flag in hive that shows the data as "new",
> you can use just a SELECT statement to get the information
>
> In general you'd use merge with parameters for that:
>
> MERGE (n:Label {id:{id}}) ON CREATE SET n.foo = {foo}, n.bar = {bar}
>
> or if you always want to update properties
>
> MERGE (n:Label {id:{id}})
> SET n.foo = {foo}, n.bar = {bar}
>
> For the actual run, there are several, options
>
> You can also export the select results to CSV (or make that CSV available via
> http) and use LOAD CSV
>
> LOAD CSV WITH HEADERS FROM "URL" as row
> MERGE (n:Label {id:row.id <http://row.id/>}) ON CREATE SET n.foo = row.foo,
> n.bar = row.bar
>
> or even
>
> LOAD CSV WITH HEADERS FROM "URL" as row
> MERGE (n:Label {id:row.id <http://row.id/>}) ON CREATE SET n += row
>
>
> Or pass in all rows of the batch in as parameters, e.g. {id:id, data:
> {col1:value, col2:value})
>
> UNWIND {rows} as row
> MERGE (n:Label {id:row.id <http://row.id/>}) ON CREATE SET n += row.data
>
> Michael
>
>> Am 27.01.2016 um 06:55 schrieb Pranab Banerjee <[email protected]
>> <mailto:[email protected]>>:
>>
>> Hi
>>
>> This is regarding incremental Data Ingestion (Modify/New-Add) to Neo4j from
>> Hive Data source.
>>
>> We need to incorporate the On-going "ChangeOnly/New" data load from source
>> table (in Hive) to Neo4j DB.
>>
>> 1. If node already exist in Neo4j DB then only update/modify that
>> specific node data.
>> 2. If node doesn't exist in Neo4j DB then append that specific node
>> data as a new
>>
>> Can you please suggest any effective solution when the data volume is at
>> scale (~5 million rows per day).
>>
>> Thanks
>> Pranab
>>
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "Neo4j" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected]
>> <mailto:[email protected]>.
>> For more options, visit https://groups.google.com/d/optout
>> <https://groups.google.com/d/optout>.
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "Neo4j" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected]
> <mailto:[email protected]>.
> For more options, visit https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>.
>
--
You received this message because you are subscribed to the Google Groups
"Neo4j" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
