Hi all, How do I link the output to an asset to the input of a graph that I will later make into an asset using AssetsDefinition.from_graph? Currently I am doing this but it is not working

@asset(
    partitions_def=daily,
    ins={"load_data": AssetIn(metadata={'allow_missing_partitions': True}, partition_mapping=NDaysPartitionMapping(days=1, offset=0))}, 
)
def trailing_window(context, load_data: pandas.DataFrame) -> list:
    df = load_data
    tidied_raw_data = tidy_raw_data(database=df, cellsParallel=context.op_config['cellsParallel'], cellsSeries=context.op_config['cellsSeries'])
    split_time_data = split_by_time(data=tidied_raw_data, cut_off_time=context.op_config['cut_off_time'])
    return split_time_data

@graph(ins={"trailing_window": GraphIn()})
def event_split_graph(trailing_window):
    df = trailing_window
    return df

event_split = AssetsDefinition.from_graph(graph_def=event_split_graph, partitions_def=daily,)

I am currently receiving this error:

@graph 'event_split_graph' returned problematic value of type <class 'dagster._core.definitions.composition.InputMappingNode'>

Hi Alessandro. A

@graph

functions as a wrapper over a set of ops with defined dependencies between them. Dagster will flatten each graph into a flat map of ops and then pass each input of the graph to the corresponding op inputs. This means that a graph must contain at least one op, and each input of the graph must be mapped to at least one op within the graph. We could do a better job of clarifying this error for sure, since it's pretty unclear what the issue is from looking at the error message.

This fixed the problem! thank you so much

This thread was really helpful. I have been struggling with this issue for a few days know trying to figure out from the docs what I am missing. @claire’s snippet was what I needed. Thanks!