How to run a workflow created in .ymal

[mursaleen-bkk]

I have question that how we can run the workflow created by us in .yaml file.
For example I have the following workflow:

name: visualize_crop_segmentation
sources:
  user_input:
    - s2.user_input
    - spaceeye.user_input
    - cdl.user_input
sinks:
  s2_rasters: s2.raster ### NEW SINK FOR VISUALIZATION
  s2_masks: s2.mask ### NEW SINK FOR VISUALIZATION
  se_rasters: spaceeye.raster ### NEW SINK FOR VISUALIZATION
  ndvi: ndvi.index_raster
  cdl: cdl.raster
parameters:
  pc_key:
tasks:
  s2: ### NEW TASK FOR VISUALIZATION
    workflow: data_ingestion/sentinel2/preprocess_s2_improved_masks
  spaceeye:
    workflow: data_ingestion/spaceeye/spaceeye_interpolation
    parameters:
      pc_key: "@from(pc_key)"
  ndvi:
    workflow: data_processing/index/index
    parameters:
      index: ndvi
  cdl:
    workflow: data_ingestion/cdl/download_cdl
edges:
  - origin: spaceeye.raster
    destination:
      - ndvi.raster

so, how i will run this workflow?

[rafaspadilha]

Hi, @mursaleen-bkk.

You can run this workflow by loading the yaml file and passing it directly as input to client.run.
You can find an example of this in the crop segmentation's visualization notebook.

client = get_default_vibe_client()

input_geometry_path = "./input_region.wkt"
time_range = (datetime(2020, 1, 1), datetime(2020, 12, 31))

# Reading the geometry file 
with open(input_geometry_path) as f:
    geometry = wkt.load(f)

# Reading the workflow
with open("wf_viz_crop_seg.yaml") as f:
    wf_dict = yaml.safe_load(f)

wf_run = client.run(wf_dict, "Crop Segmentation Workflow Visualization", geometry=geometry, time_range=time_range)
wf_run.monitor()

In this case, wf_viz_crop_seg.yaml is a custom workflow similar to yours.