Library langchain.graphs is returning only 50 rows max. version 0.0.292

Integrations
, ,
1

Hi everyone,

I´m recently using LagChain to do some text embeddigns , but the class Neo4jGraph is just returning 50 rows when my graph model have around 14,000 nodes.

Here is some tests

from langchain.graphs import Neo4jGraph
graph = Neo4jGraph(
    url=NEO4J_URI,
    username=NEO4J_USERNAME,
    password=NEO4J_PASSWORD
)
results=graph.query(""" 
MATCH (c:CONTRATO) 
RETURN c.Id_Contrato as idContrato""")
df=pd.DataFrame(results,columns=["idContrato"])
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 50 entries, 0 to 49
Data columns (total 1 columns):

Column Non-Null Count Dtype

0 idContrato 50 non-null int64
dtypes: int64(1)
memory usage: 528.0 bytes

Using this is OK:

from graphdatascience import GraphDataScience
gds = GraphDataScience(NEO4J_URI, auth=(NEO4J_USERNAME, NEO4J_PASSWORD))
results=gds.run_cypher("""MATCH (c:CONTRATO) RETURN c.Id_Contrato as idContrato """)
df2=pd.DataFrame(results,columns=["idContrato"])
df2.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 14494 entries, 0 to 14493
Data columns (total 1 columns):

Column Non-Null Count Dtype

0 idContrato 14494 non-null int64
dtypes: int64(1)
memory usage: 113.4 KB

There´s someone in the team of Neo4j participate with LangChain to help me?

Thanks in advanced

2

It seems that I have to divide ther work in two fases :

It seems that I have to divide the work in two fases
1.- the first one using import Graph Data Science to returning all rows that I need, and then process the embeddings and create de vectors
2.- The second using the langchain.graphs only for the Vector Search

** anyway the OpenAI emmbeding API Its getting trouble with a big batch request