@Scout Can I use .map in a graph_asset?

The documentation does not explicitly mention the use of

.map

within a

graph_asset

. The

graph_asset

decorator is used for defining an asset that's computed using a graph of ops, and the dependencies between the ops are specified inside the body of the decorated function. However, the use of

.map

specifically within this context is not detailed in the documentation provided. If you're looking to apply an operation to multiple elements or perform parallel execution within a

graph_asset

, you might need to look into other patterns supported by Dagster, such as dynamic outputs or fan-in/fan-out patterns, which are mentioned in the context of op graphs. For a definitive answer or example usage, you may want to refer to the latest Dagster documentation or explore the Dagster community resources such as GitHub discussions or issues for any updates or community practices regarding the use of

.map

in

graph_asset

.

@Scout Does graph_asset requires a return value?

Yes, when using the

@graph_asset

decorator in Dagster, you are expected to return a value from the decorated function. This return value represents the asset or assets that are computed by the graph of ops within the

graph_asset

. The return value can be a single output or a dictionary mapping output names to return values from invoked nodes. Here's an example from the documentation that illustrates a

graph_asset

with a return value:

python @graph_asset def my_graph_assets():     bar_1, bar_2 = bar()     foo_1, foo_2 = foo(bar_1)     return {"foo_asset": foo_1, "baz_asset": baz(foo_2, bar_2)}

In this example,

my_graph_assets

is a

graph_asset

that invokes two ops,

bar

and

foo

, and then returns a dictionary mapping the names of the assets to their respective values. The keys in the dictionary (

"foo_asset"

and

"baz_asset"

) correspond to the names of the assets being defined, and the values are the outputs from the ops that compute the assets. If you attempt to define a

graph_asset

without returning a value, or if the return value does not match the expected outputs, Dagster will raise an error indicating that the definition is invalid.