enh: change measurement identifier upon filtered export

CHANGELOG

@@ -7,6 +7,7 @@
  - enh: more caching of event size and shape for HDF5 format
  - enh: faster computation of contour length for DCOR format
  - enh: use dcserv version 2 in DCOR format (fast S3 access)
+ - enh: change measurement identifier upon filtered export
  - setup: pin s3fs>=2023.10.0
  - setup: pin upper bounds of dependencies
 0.54.2

dclab/rtdc_dataset/export.py

@@ -1,6 +1,7 @@
 """Export RT-DC measurement data"""
 import codecs
 import pathlib
+import uuid
 import warnings
 
 import h5py
@@ -240,6 +241,12 @@ def hdf5(self, path, features=None, filtered=True,
         # add user-defined metadata
         if "user" in self.rtdc_ds.config:
             meta["user"] = self.rtdc_ds.config["user"].copy()
+        if filtered:
+            # Define a new measurement identifier, so that we are not running
+            # into any problems with basins being defined for filtered data.
+            ds_run_id = self.rtdc_ds.get_measurement_identifier()
+            random_ap = str(uuid.uuid4())[:4]
+            meta["experiment"]["run identifier"] = f"{ds_run_id}-{random_ap}"
 
         if filtered:
             filtarr = self.rtdc_ds.filter.all

tests/test_export.py

@@ -312,6 +312,49 @@ def test_hdf5_filtered_index():
     assert ds2.config["experiment"]["event count"] == n - 1
 
 
+@pytest.mark.filterwarnings(
+    "ignore::dclab.rtdc_dataset.config.WrongConfigurationTypeWarning")
+def test_hdf5_identifier_different():
+    """Make sure the identifier is changed upon export"""
+    pin = retrieve_data("fmt-hdf5_image-bg_2020.zip")
+    pout = pin.with_name("exported.rtdc")
+    with new_dataset(pin) as din:
+        mid_in = din.get_measurement_identifier()
+        assert len(din) == 5
+        assert mid_in
+        din.config["filtering"]["deform min"] = np.mean(din["deform"])
+        din.config["filtering"]["deform max"] = np.max(din["deform"])
+        din.apply_filter()
+        din.export.hdf5(pout, ["area_um", "deform"], filtered=True)
+
+    with new_dataset(pout) as dout:
+        mid_out = dout.get_measurement_identifier()
+        assert mid_out.startswith(mid_in)
+        assert mid_out != mid_in
+        assert len(dout) == 3
+
+
+@pytest.mark.filterwarnings(
+    "ignore::dclab.rtdc_dataset.config.WrongConfigurationTypeWarning")
+def test_hdf5_identifier_same_unfiltered():
+    """Make sure the identifier is changed upon export"""
+    pin = retrieve_data("fmt-hdf5_image-bg_2020.zip")
+    pout = pin.with_name("exported.rtdc")
+    with new_dataset(pin) as din:
+        mid_in = din.get_measurement_identifier()
+        assert len(din) == 5
+        assert mid_in
+        din.config["filtering"]["deform min"] = np.mean(din["deform"])
+        din.config["filtering"]["deform max"] = np.max(din["deform"])
+        din.apply_filter()
+        din.export.hdf5(pout, ["area_um", "deform"], filtered=False)
+
+    with new_dataset(pout) as dout:
+        mid_out = dout.get_measurement_identifier()
+        assert mid_out == mid_in
+        assert len(dout) == 5
+
+
 def test_hdf5_image_bg():
     n = 65
     keys = ["image", "image_bg"]
@@ -618,11 +661,3 @@ def test_tsv_not_filtered():
     edest = tempfile.mkdtemp()
     f1 = join(edest, "test.tsv")
     ds.export.tsv(f1, keys, filtered=False)
-
-
-if __name__ == "__main__":
-    # Run all tests
-    _loc = locals()
-    for _key in list(_loc.keys()):
-        if _key.startswith("test_") and hasattr(_loc[_key], "__call__"):
-            _loc[_key]()