variability map: avoid float output, fix band name

Author: jdries

algorithm_catalog/vito/variabilitymap/openeo_udp/generate.py

@@ -38,14 +38,14 @@ def get_variabilitymap(
 
     biopar_type = "FAPAR"
 
-    mask = mask = scl.process("to_scl_dilation_mask", data=scl)
+    mask = scl.process("to_scl_dilation_mask", data=scl)
     S2_bands_mask = s2_cube.mask(mask)
 
     # fetch udf to reduce bands
     reduce_bands_udf = openeo.UDF.from_file("shub_fapar_udf.py")
     S2_bands_mask_reduced = S2_bands_mask.reduce_bands(reduce_bands_udf)
 
-    input_data = S2_bands_mask_reduced.add_dimension(label=biopar_type, name=biopar_type, type='bands')
+    input_data = S2_bands_mask_reduced.add_dimension(label=biopar_type, name='bands', type='bands')
 
     ################ get and apply variability map udf #################
 
@@ -57,7 +57,9 @@ def get_variabilitymap(
                                             runtime='Python', context={
             'mask_value': mask_value, 'raw': {"from_parameter": "raw"}, 'band': biopar_type
         })
-    variabilitymap = input_data.apply_polygon(geometries=spatial_extent, process=udf_process, mask_value=mask_value)
+    variabilitymap = (input_data.apply_polygon(geometries=spatial_extent, process=udf_process, mask_value=mask_value)
+                      .linear_scale_range(0,100,0,100).rename_labels(dimension= 'bands', target=['variability']))
+
     return variabilitymap
 
 

algorithm_catalog/vito/variabilitymap/openeo_udp/variabilitymap.json

@@ -71,7 +71,7 @@
                   "from_parameter": "data"
                 },
                 "runtime": "Python",
-                "udf": "\"\"\"\nImport the biopar library and the BioParNp class to calculate the FAPAR index.\nIt is the Python implementation of biophyscial parameter computation, \nas described here: http://step.esa.int/docs/extra/ATBD_S2ToolBox_L2B_V1.1.pdf\n\"\"\"\nfrom functools import lru_cache\nimport numpy as np\nfrom typing import Dict\nfrom numpy import cos, radians\n\nfrom openeo.udf.xarraydatacube import XarrayDataCube\nfrom biopar.bioparnp import BioParNp\n\n\n@lru_cache(maxsize=6)\ndef get_bioparrun(biopar) -> BioParNp:\n    return BioParNp(version='3band', parameter=biopar, singleConfig = True)\n \n\ndef apply_datacube(cube: XarrayDataCube, context: Dict) -> XarrayDataCube:\n    ds_date = cube.get_array()\n\n    ### LOAD THE DIFFERENT REQUIRED BANDS FOR THE 8-BAND FAPAR\n    scaling_bands = 0.0001\n\n    saa = ds_date.sel(bands='sunAzimuthAngles')\n    sza = ds_date.sel(bands=\"sunZenithAngles\")\n    vaa = ds_date.sel(bands=\"viewAzimuthMean\")\n    vza = ds_date.sel(bands=\"viewZenithMean\")\n\n    B03 = ds_date.sel(bands='B03') * scaling_bands\n    B04 = ds_date.sel(bands='B04') * scaling_bands\n    B8 = ds_date.sel(bands='B08') * scaling_bands\n\n    g1 = cos(radians(vza))\n    g2 = cos(radians(sza))\n    g3 = cos(radians(saa - vaa))\n\n    #### FLATTEN THE ARRAY ####\n    flat = list(map(lambda arr: arr.flatten(),\n                    [B03.values, B04.values, B8.values, g1.values, g2.values, g3.values]))\n    bands = np.array(flat)\n\n    #### CALCULATE THE BIOPAR BASED ON THE BANDS #####\n    \n    image = get_bioparrun('FAPAR').run(bands,\n                                       output_scale=1,\n                                       output_dtype=np.float32,\n                                       minmax_flagging=False)  # netcdf algorithm\n    as_image = image.reshape((g1.shape))\n\n    ## SET NOTDATA TO NAN\n    as_image[np.where(np.isnan(B03))] = np.nan\n    xr_biopar = vza.copy()\n    xr_biopar.values = as_image\n    return XarrayDataCube(xr_biopar)"
+                "udf": "\"\"\"\nImport the biopar library and the BioParNp class to calculate the FAPAR index.\nIt is the Python implementation of biophysical parameter computation, \nas described here: http://step.esa.int/docs/extra/ATBD_S2ToolBox_L2B_V1.1.pdf\n\"\"\"\nfrom functools import lru_cache\nimport numpy as np\nfrom typing import Dict\nfrom numpy import cos, radians\n\nfrom openeo.udf.xarraydatacube import XarrayDataCube\nfrom biopar.bioparnp import BioParNp\n\n\n@lru_cache(maxsize=6)\ndef get_bioparrun(biopar) -> BioParNp:\n    return BioParNp(version='3band', parameter=biopar, singleConfig = True)\n \n\ndef apply_datacube(cube: XarrayDataCube, context: Dict) -> XarrayDataCube:\n    ds_date = cube.get_array()\n\n    ### LOAD THE DIFFERENT REQUIRED BANDS FOR THE 8-BAND FAPAR\n    scaling_bands = 0.0001\n\n    saa = ds_date.sel(bands='sunAzimuthAngles')\n    sza = ds_date.sel(bands=\"sunZenithAngles\")\n    vaa = ds_date.sel(bands=\"viewAzimuthMean\")\n    vza = ds_date.sel(bands=\"viewZenithMean\")\n\n    B03 = ds_date.sel(bands='B03') * scaling_bands\n    B04 = ds_date.sel(bands='B04') * scaling_bands\n    B8 = ds_date.sel(bands='B08') * scaling_bands\n\n    g1 = cos(radians(vza))\n    g2 = cos(radians(sza))\n    g3 = cos(radians(saa - vaa))\n\n    #### FLATTEN THE ARRAY ####\n    flat = list(map(lambda arr: arr.flatten(),\n                    [B03.values, B04.values, B8.values, g1.values, g2.values, g3.values]))\n    bands = np.array(flat)\n\n    #### CALCULATE THE BIOPAR BASED ON THE BANDS #####\n    \n    image = get_bioparrun('FAPAR').run(bands,\n                                       output_scale=1,\n                                       output_dtype=np.float32,\n                                       minmax_flagging=False)  # netcdf algorithm\n    as_image = image.reshape((g1.shape))\n\n    ## SET NOTDATA TO NAN\n    as_image[np.where(np.isnan(B03))] = np.nan\n    xr_biopar = vza.copy()\n    xr_biopar.values = as_image\n    return XarrayDataCube(xr_biopar)"
               },
               "result": true
             }
@@ -86,7 +86,7 @@
           "from_node": "reducedimension1"
         },
         "label": "FAPAR",
-        "name": "FAPAR",
+        "name": "bands",
         "type": "bands"
       }
     },
@@ -122,6 +122,43 @@
             }
           }
         }
+      }
+    },
+    "apply1": {
+      "process_id": "apply",
+      "arguments": {
+        "data": {
+          "from_node": "applypolygon1"
+        },
+        "process": {
+          "process_graph": {
+            "linearscalerange1": {
+              "process_id": "linear_scale_range",
+              "arguments": {
+                "inputMax": 100,
+                "inputMin": 0,
+                "outputMax": 100,
+                "outputMin": 0,
+                "x": {
+                  "from_parameter": "x"
+                }
+              },
+              "result": true
+            }
+          }
+        }
+      }
+    },
+    "renamelabels1": {
+      "process_id": "rename_labels",
+      "arguments": {
+        "data": {
+          "from_node": "apply1"
+        },
+        "dimension": "bands",
+        "target": [
+          "variability"
+        ]
       },
       "result": true
     }