Hi all! Is is possible to use `graph_asset`s with single-run backfills? I tried calling

context.asset_partitions_time_window

inside an

op

which is part of my

graph_asset

, but got an error. Code to reproduce:

from dagster import (
    DailyPartitionsDefinition,
    Definitions,
    OpExecutionContext,
    asset,
    graph_asset,
    op,
)

day_partitions_def = DailyPartitionsDefinition(start_date="2023-05-01")


@op
def extract(context: OpExecutionContext):
    # Error happens here!
    tw = context.asset_partitions_time_window_for_output()

    <http://context.log.info|context.log.info>(f"Extracting Data: {tw.start} - {tw.end}")
    x = 1
    <http://context.log.info|context.log.info>(f"Extracted Data: {x}")
    return x


@op
def transform(context: OpExecutionContext, x):
    <http://context.log.info|context.log.info>(f"Transform Input: {x}")
    y = x + 1
    <http://context.log.info|context.log.info>(f"Transform Output: {y}")
    return y


@op
def load(context: OpExecutionContext, y):
    <http://context.log.info|context.log.info>("Loaded Data: {y}")


@graph_asset(partitions_def=day_partitions_def)
def my_graph_asset():
    x = extract()
    y = transform(x)
    return load(y


defs = Definitions(
    assets=[my_graph_asset],
)

I found a thread where @owen suggests using

context.partition_key

inside the

op

, but it does not work with the single-run backfill.