The basic unit of computation in Dagster is the op. In certain cases it is desirable to run the same op multiple times on different pieces of similar data.
Dynamic outputs are the tool Dagster provides to allow resolving the pieces of data at runtime and having downstream copies of the ops created for each piece.
While, the implementation of expensive_computation can internally do something to parallelize the compute, if anything goes wrong with any part we have to restart the whole computation.
With this motivation we will break up the computation using Dynamic Outputs. First we will define our new op that will use dynamic outputs. First we use DynamicOut to declare that this op will return dynamic outputs. Then in the function we yield a number of DynamicOutput objects that each contain a value and a unique mapping_key.
@op(out=DynamicOut())defload_pieces():
large_data = load_big_data()for idx, piece in large_data.chunk():yield DynamicOutput(piece, mapping_key=idx)
Then after creating ops for our downstream operations, we can put them all together in a job.
Within our @job decorated composition function, the object representing the dynamic output can not be passed directly to another op. Either map or collect must be invoked on it.
map takes a Callable which receives a single argument. This callable is evaluated once, and any invoked op that is passed the input argument will establish dependencies. The ops downstream of a dynamic output will be cloned for each dynamic output, and identified using the associated mapping_key. The return value from the callable is captured and wrapped in an object that allows for subsequent map or collect calls.
collect creates a fan-in dependency over all the dynamic copies. The dependent op will receive a list containing all the values.
In addition to yielding, DynamicOutput objects can also be returned as part of a list.
from dagster import DynamicOut, DynamicOutput, op
from typing import List
@op(out=DynamicOut())defreturn_dynamic()-> List[DynamicOutput[str]]:
outputs =[]for idx, page_key in get_pages():
outputs.append(DynamicOutput(page_key, mapping_key=idx))return outputs
DynamicOutput can be used as a generic type annotation describing the expected type of the output.
Multiple outputs are returned via a namedtuple, where each entry can be used via map or collect.
@op(
out={"values": DynamicOut(),"negatives": DynamicOut(),},)defmultiple_dynamic_values():for i inrange(2):yield DynamicOutput(i, output_name="values", mapping_key=f"num_{i}")yield DynamicOutput(-i, output_name="negatives", mapping_key=f"neg_{i}")@jobdefmultiple():# can unpack on assignment (order based)
values, negatives = multiple_dynamic_values()
process(values.collect())
process(negatives.map(echo).collect())# can use map or collect as usual# or access by name
outs = multiple_dynamic_values()
process(outs.values.collect())
process(outs.negatives.map(echo).collect())