Give me an example of a graph of ops that uses the output of an upstream op as the graph input

Certainly! Here's an example of a graph where the output of an upstream op is used as the input to the graph. This example is based on the concepts found in the documentation:

python from dagster import op, graph, In, Out  @op def upstream_op():     # This op would perform some computation and return a result     return "result_from_upstream"  @op def process_data(data: str):     # This op takes the result from the upstream op and does further processing     return f"processed_{data}"  @graph def my_graph(input_data: str):     processed_data = process_data(input_data)     return processed_data  # Now, to use the output of `upstream_op` as the input to `my_graph`, you would # execute the graph and pass the output of `upstream_op` to it. result = my_graph(upstream_op())

In this example,

upstream_op

is an op that generates some data. The

my_graph

graph takes an input, which is expected to be the output from

upstream_op

. Inside

my_graph

, the

process_data

op is called with the input data, which processes it further. Please note that this example is for illustrative purposes to show how the output of an upstream op can be used as an input to a graph. In a real-world scenario, you would likely have more complex logic within your ops and graphs, and you would execute the graph within a job or a test environment.