Start end-to-end transfer method to ilastik object model

extensions/chaeo/examples/label_patches.py

@@ -13,3 +13,5 @@ if __name__ == '__main__':
         where_output=autonumber_new_directory(root, 'labeled_patches'),
     )
     print('Finished')
+
+

extensions/chaeo/examples/transfer_labels_to_ilastik_object_classifier.py

@@ -4,8 +4,22 @@ import h5py
 import ilastik.applets.objectClassification
 import numpy as np
 import pandas as pd
+import tifffile
 
 from extensions.ilastik.models import IlastikObjectClassifierModel
+from model_server.accessors import generate_file_accessor
+
+def get_dataset_info(h5):
+    lane = 'Input Data/infos/lane0000'
+    info = {}
+    for gk in ['Raw Data', 'Segmentation Image']:
+        info[gk] = {}
+        for dk in ['location', 'filePath', 'shape']:
+            try:
+                info[gk][dk] = h5[f'{lane}/{gk}/{dk}'][()]
+            except Exception as e:
+                print(e)
+    return info
 
 def transfer_labels_to_ilastik_ilp(ilp, df_stack_meta):
 
@@ -22,6 +36,9 @@ def transfer_labels_to_ilastik_ilp(ilp, df_stack_meta):
                 hf.write(f'{line[0]} --- {line[1]}\n')
         h5.visititems(lambda k, v: print(k + ' : ' + str(v)))
 
+        # put certain h5 groups in scope
+        h5info = get_dataset_info(h5)
+
         # change key of label names
         ln = ['none'] + list(df_stack_meta.sort_values('annotation_class_id').annotation_class.unique())
         del h5['ObjectClassification/LabelNames']
@@ -39,9 +56,44 @@ def transfer_labels_to_ilastik_ilp(ilp, df_stack_meta):
             ds[1] = float(df_stack_meta.loc[df_stack_meta.zi == idx, 'annotation_class_id'].iat[0])
             print(f'Changed label {ti} from {la_old} to {ds[1]}')
 
+def generate_ilastik_object_classifier(template_ilp, where_training: str):
+
+    # validate input data
+    where = Path(where_training)
+    zstacks = {
+        'raw': {
+            'path': where / 'zstack_train_raw.tif',
+        },
+        'seg': {
+            'path': where / 'zstack_train_mask.tif',
+        }
+    }
+
+    for k, v in zstacks.items():
+        # assert v['path'].exists(), 'Could not find input z-stack: ' + v['path']
+        # ff = tifffile.imread(v['path'])
+        # v['nz'] = ff.shape[0]
+        # v['hw'] = ff.shape[1:2]
+        # v['dtype'] = ff.dtype
+        v['acc'] = generate_file_accessor(v['path'])
+    assert zstacks['raw']['acc'].is_binary()
+
+    assert zstacks['raw']
+
+    with h5py.File(template_ilp, 'r+') as h5:
+        pass
+
+
 if __name__ == '__main__':
-    ilp = 'c:/Users/rhodes/model-server/ilastik/test_autolabel_obj.ilp'
+    ilp = 'c:/Users/rhodes/model-server/ilastik/test_autolabel_obj - Copy.ilp'
+    # ilp = 'c:/Users/rhodes/model-server/ilastik/test_template_obj.ilp'
+
     df = pd.read_csv(
         'c:/Users/rhodes/projects/proj0011-plankton-seg/exp0009/output/labeled_patches-20231014-0002/train_stack.csv'
     )
-    transfer_labels_to_ilastik_ilp(ilp, df)
\ No newline at end of file
+    # transfer_labels_to_ilastik_ilp(ilp, df)
+
+    generate_ilastik_object_classifier(
+        ilp,
+        'c:/Users/rhodes/projects/proj0011-plankton-seg/exp0009/output/labeled_patches-20231014-0002'
+    )
\ No newline at end of file