Simple example of Multipleresults #65
Conversation
There was a problem hiding this comment.
I would not keep both examples, and simply use the one where mutiple results are saved.
It would be nice if you could include an image of the process graph prior to execution and provide some details on how this graph now contains multiple saving nodes which are all processed and saved at once
Included the feature in same example.
Updated |
| @@ -27,7 +27,7 @@ The `Demonstrates` column summarizes the key openEO functionality used in each c | |||
| | [WorldCereal](./WorldCereal/) | `Jupyter` | WorldCereal data extraction sample. | openEO processes `merge_cubes`, loading **WorldCereal** data | | |||
| | [RVI](./RVI/) | `Jupyter` | Calculate Radar Vegetation Index | openEO processes `sar_backscatter`, `spectral_nidices.compute_indices`; **plotting** mean result and timeseries; **Awesome Spectral Indices** | | |||
| | [OilSpill](./OilSpill/) | `Jupyter` | Oil Spill mapping with Sentinel-1 layer. | openEO processes `sar_backscatter`, `apply`, `apply_kernel`, `rename_labels`, `merge_cubes`; **plotting** binary image | | |||
| | [ForestFire](./ForestFire/) | `Jupyter` | Wildfire mapping using Sentinel-2 | openEO processes `apply_kernel`,`ndvi` `spectral_nidices.compute_indices`; **plotting** comparative visualisation; **Awesome Spectral Indices** | | |||
| | [ForestFire](./ForestFire/) | `Jupyter` | Wildfire mapping using Sentinel-2 | openEO processes `apply_kernel`,`ndvi`,`spectral_nidices.compute_indices`,`MultiResult`; **plotting** comparative visualisation; **Awesome Spectral Indices** | | |||
There was a problem hiding this comment.
I would reconsider this title, ideally it hints towards both the application as the openeo functionality used (saving multiple results)
"Forest Fire Analysis: Saving Multiple Results in One Job" or so?
There was a problem hiding this comment.
In general the notebook is quite convoluted, please take a closer look from story telling perspective on how you can impove it. I would definetely make some choices and try to shorten it.
the introduction is not coherent, and when the process graph is shown, I do not believe we discuss the multiple save result nodes?
There is also some code replication going on with the load collection, it could help to wrap this in order to avoid repeating yourself
"Instead of executing on the backend immediately to get the ndvi_pre as an output, let us save it as an intermediate result with a specific name and format for the moment." I would rephrase towars something like: "since we are interested in this intermediant result, we attach a save result ....."
Personally I find
(i) NRT mode and
(ii) Post-event mode.
a bit confusing, do we need both?
#51
Similar task is also in progress with: #64