Basic API

To begin your querying session, get a connection to the EvaDB using connect:

import evadb

from eva.interfaces.relational.db import EVAConnection, connect
conn = connect()

You can then use this connection to run queries:

conn.load("online_video.mp4", "youtube_video", "video").execute()
conn.query("CREATE TABLE IF NOT EXISTS youtube_video_text AS SELECT SpeechRecognizer(audio) FROM youtube_video;").execute()

Warning

It is important to call execute to run the actual query.

EvaDB uses a lazy query execution technique to improve performance. Calling conn.query("...") will only construct and not run the query. Calling conn.query("...").execute() will both construct and run the query.

EVAConnection Interface

load(file_regex, table_name, format, **kwargs)	Load data from files into a table in the database.
query(sql_query)	Execute a SQL query.
table(table_name)	Get a EVARelation object representing a table in the database.

EVARelation Interface

select(expr)	Projects a set of expressions and returns a new EVARelation.
execute()	Transform the relation into a result set
cross_apply(expr, alias)	Execute a expr on all the rows of the relation
filter(expr)	Filters rows using the given condition.
df()	Execute and fetch all rows as a pandas DataFrame

Advanced API

EVAConnection Interface

create_vector_index(index_name, table_name, ...)	Create a vector index on a table in the database.
df()	Get the result of the last executed query on the connection as a pandas DataFrame.

EVARelation Interface

alias(alias)	Returns a new Relation with an alias set.
limit(num)	Limits the result count to the number specified.
order(order_expr)	Reorder the relation based on the order_expr
show()	Execute and fetch all rows as a pandas DataFrame
sql_query()	Get the SQL query that is equivalent to the relation