Replies: 4 comments 9 replies
-
|
Some quick comments: the initial graph is invalid I think: {
"apply1": {
"arguments": {
"data": {"from_node": "load1"},
"dimension": "t",
"process": {
"process_graph": {
"max1": {
"arguments": {"data": {"from_parameter": "data"}},
"process_id": "max",
"result": true
}
}
}
},
"process_id": "apply_dimension"
}apply_dimensions expects a process that takes an array as input and produces again an array. Here the output (of In the new process graph I also see "array1": {
"process_id": "array_create",
"arguments": {
"data": [
{"from_node": "min1" },
{"from_node": "max1" },
{"from_node": "sd1" },
... |
Beta Was this translation helpful? Give feedback.
-
|
Indeed, the initial UDP is invalid although it seems to work on VITO. And you are correct with array_create, that's even better! I'll update my proposal above. |
Beta Was this translation helpful? Give feedback.
-
I guess we implicitly/automatically wrap the scalar in an array here. We better throw an error to play well by the openEO API rules? |
Beta Was this translation helpful? Give feedback.
-
|
In theory, yes. In practice it could break some existing use cases that exploit the issue. Maybe starting with a warning might be better? |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
And, if I understand correctly, the |
Beta Was this translation helpful? Give feedback.
-
|
Yes, I think so. It's flatten_dimensions though. |
Beta Was this translation helpful? Give feedback.
-
|
flatten_dimensions indeed. Typing too fast |
Beta Was this translation helpful? Give feedback.
-
|
FYI, instead of flatten_dimension, in our other online example we still set target='bands' in apply_dimension If that isn't correct, we should flag it as such. |
Beta Was this translation helpful? Give feedback.
-
|
What happens in this case is at least undocumented in apply_dimension. I think we only used this as a workaround and as such it was never added in apply_dimension?! So I think we should update that if possible now. |
Beta Was this translation helpful? Give feedback.
-
|
So is there a recommended implementation for temporal metrics as discussed above? Should we follow the VITO example above? That one works, but for capacity building purposes I don't find it very intuitive or "opene-ophonic".. |
Beta Was this translation helpful? Give feedback.
-
|
I'd say: |
Beta Was this translation helpful? Give feedback.
-
|
Apologies, don't have time to give a proper answer, but the short one is that we can't do flatten_dimensions and need to properly discuss after the BiDS preparation. |
Beta Was this translation helpful? Give feedback.
-
I stumbled across a user-defined process (see the "Original" UDP below) that computes S1 SAR metrics:
I found that it looks overly difficult and verbose. I was wondering whether there's a simpler (and probably more performant) way to achieve this in openEO. I think there is:
See the "New" UDP below. It should lead to exactly the same results. The only issue that I see is that in the underlying back-end (openEO Platform via VITO) the
flatten_dimensionprocess is missing (and you may want to rename the target dimension frommetricstobandsto get exactly the same result).Anyone has any thoughts? @soxofaan
Original
{ "process_graph": { "12": { "arguments": { "cube1": { "from_node": "merge11" }, "cube2": { "from_node": "merge13" } }, "process_id": "merge_cubes" }, "apply1": { "arguments": { "data": { "from_node": "load1" }, "dimension": "t", "process": { "process_graph": { "max1": { "arguments": { "data": { "from_parameter": "data" } }, "process_id": "max", "result": true } } } }, "process_id": "apply_dimension" }, "apply2": { "arguments": { "data": { "from_node": "load1" }, "dimension": "t", "process": { "process_graph": { "min1": { "arguments": { "data": { "from_parameter": "data" } }, "process_id": "min", "result": true } } } }, "process_id": "apply_dimension" }, "apply3": { "arguments": { "data": { "from_node": "load1" }, "dimension": "t", "process": { "process_graph": { "sd1": { "arguments": { "data": { "from_parameter": "data" } }, "process_id": "sd", "result": true } } } }, "process_id": "apply_dimension" }, "apply4": { "arguments": { "data": { "from_node": "load1" }, "dimension": "t", "process": { "process_graph": { "quantiles1": { "arguments": { "data": { "from_parameter": "data" }, "probabilities": [ 0.1 ] }, "process_id": "quantiles", "result": true } } } }, "process_id": "apply_dimension" }, "apply8": { "arguments": { "data": { "from_node": "load1" }, "dimension": "t", "process": { "process_graph": { "quantiles1": { "arguments": { "data": { "from_parameter": "data" }, "probabilities": [ 0.9 ] }, "process_id": "quantiles", "result": true } } } }, "process_id": "apply_dimension" }, "apply9": { "arguments": { "data": { "from_node": "load1" }, "dimension": "t", "process": { "process_graph": { "quantiles1": { "arguments": { "data": { "from_parameter": "data" }, "probabilities": [ 0.5 ] }, "process_id": "quantiles", "result": true } } } }, "process_id": "apply_dimension" }, "load1": { "arguments": { "bands": [ "VH", "VV" ], "id": "S1_GRD_SIGMA0_ASCENDING", "spatial_extent": { "east": 9.03217171637845, "north": 44.45184992322663, "south": 44.38021695542935, "west": 8.855289792705683 }, "temporal_extent": [ "2023-05-01T00:00:00Z", "2023-05-31T00:00:00Z" ] }, "process_id": "load_collection" }, "merge11": { "arguments": { "cube1": { "from_node": "rename7" }, "cube2": { "from_node": "rename6" } }, "process_id": "merge_cubes" }, "merge12": { "arguments": { "cube1": { "from_node": "rename16" }, "cube2": { "from_node": "rename15" } }, "process_id": "merge_cubes" }, "merge13": { "arguments": { "cube1": { "from_node": "rename5" }, "cube2": { "from_node": "rename10" } }, "process_id": "merge_cubes" }, "merge14": { "arguments": { "cube1": { "from_node": "merge12" }, "cube2": { "from_node": "12" } }, "process_id": "merge_cubes" }, "rename10": { "arguments": { "data": { "from_node": "apply9" }, "dimension": "bands", "source": [ "VH", "VV" ], "target": [ "VH_Q50", "VV_Q50" ] }, "process_id": "rename_labels" }, "rename15": { "arguments": { "data": { "from_node": "apply4" }, "dimension": "bands", "source": [ "VH", "VV" ], "target": [ "VH_Q10", "VV_Q10" ] }, "process_id": "rename_labels" }, "rename16": { "arguments": { "data": { "from_node": "apply3" }, "dimension": "bands", "source": [ "VH", "VV" ], "target": [ "VH_SD", "VV_SD" ] }, "process_id": "rename_labels" }, "rename5": { "arguments": { "data": { "from_node": "apply8" }, "dimension": "bands", "source": [ "VH", "VV" ], "target": [ "VH_Q90", "VV_Q90" ] }, "process_id": "rename_labels" }, "rename6": { "arguments": { "data": { "from_node": "apply1" }, "dimension": "bands", "source": [ "VH", "VV" ], "target": [ "VH_max", "VV_max" ] }, "process_id": "rename_labels" }, "rename7": { "arguments": { "data": { "from_node": "apply2" }, "dimension": "bands", "source": [ "VH", "VV" ], "target": [ "VH_min", "VV_min" ] }, "process_id": "rename_labels" }, "save10": { "arguments": { "data": { "from_node": "merge14" }, "format": "GTIFF" }, "process_id": "save_result", "result": true } } }New
{ "process_graph": { "load1": { "process_id": "load_collection", "arguments": { "bands": [ "VH", "VV" ], "id": "S1_GRD_SIGMA0_ASCENDING", "spatial_extent": { "east": 9.03217171637845, "north": 44.45184992322663, "south": 44.38021695542935, "west": 8.855289792705683 }, "temporal_extent": [ "2023-05-01T00:00:00Z", "2023-05-31T00:00:00Z" ] } }, "apply1": { "process_id": "apply_dimension", "arguments": { "data": { "from_node": "load1" }, "dimension": "t", "process": { "process_graph": { "array1": { "process_id": "array_create", "arguments": { "data": [ { "from_node": "min1" }, { "from_node": "max1" }, { "from_node": "sd1" }, { "from_node": "quantiles01" }, { "from_node": "quantiles05" }, { "from_node": "quantiles09" } ] }, "result": true }, "max1": { "process_id": "max", "arguments": { "data": { "from_parameter": "data" } } }, "min1": { "process_id": "min", "arguments": { "data": { "from_parameter": "data" } } }, "quantiles01": { "process_id": "quantiles", "arguments": { "data": { "from_parameter": "data" }, "probabilities": [ 0.1 ] } }, "quantiles05": { "process_id": "quantiles", "arguments": { "data": { "from_parameter": "data" }, "probabilities": [ 0.5 ] } }, "quantiles09": { "process_id": "quantiles", "arguments": { "data": { "from_parameter": "data" }, "probabilities": [ 0.9 ] } }, "sd1": { "process_id": "sd", "arguments": { "data": { "from_parameter": "data" } } } } } } }, "rename1": { "process_id": "rename_labels", "arguments": { "data": { "from_node": "apply1" }, "target": [ "min", "max", "SD", "Q10", "Q50", "Q90" ], "dimension": "t" } }, "flatten1": { "process_id": "flatten_dimensions", "arguments": { "data": { "from_node": "rename1" }, "dimensions": [ "bands", "t" ], "target_dimension": "metrics", "label_separator": "_" } }, "save10": { "process_id": "save_result", "arguments": { "data": { "from_node": "flatten1" }, "format": "GTIFF" }, "result": true } } }Beta Was this translation helpful? Give feedback.
All reactions