Merge branch 'develop'

.gitignore

@@ -28,3 +28,4 @@ wheels/
 
 # Jupyter Notebook
 .ipynb_checkpoints
+*.ipynb

flowkit/_models/sample.py

@@ -154,6 +154,9 @@ def __init__(
         except KeyError:
             self.acquisition_date = None
 
+        # TODO: Allow user to set some sort of Sample ID or the orig filename,
+        #       would be useful for Samples created from data arrays or if
+        #       2 FCS files had the same file name.
         try:
             self.original_filename = self.metadata['fil']
         except KeyError:
@@ -850,15 +853,15 @@ def export(
 
         ext = os.path.splitext(filename)[-1]
 
-        if ext == 'csv':
+        if ext == '.csv':
             np.savetxt(
                 output_path,
                 events,
                 delimiter=',',
                 header=",".join(self.pnn_labels),
                 comments=''
             )
-        elif ext == 'fcs':
+        elif ext == '.fcs':
             fh = open(output_path, 'wb')
 
             flowio.create_fcs(

flowkit/_models/session.py

@@ -83,6 +83,9 @@ def gate_sample(data):
 
 
 def gate_samples(gating_strategy, samples, verbose):
+    # TODO: Looks like multiprocessing can fail for very large workloads (lots of gates), maybe due
+    #       to running out of memory. Will investigate further, but for now maybe provide an option
+    #       for turning off multiprocessing so end user can avoid this issue if it occurs.
     if multi_proc:
         if len(samples) < mp.cpu_count():
             proc_count = len(samples)
@@ -141,8 +144,12 @@ def gates(self):
 
     # start pass through methods for GatingStrategy
     def add_gate(self, gate):
+        # TODO: allow adding multiple gates at once, while still allowing a single gate. Check if list or Gate instance
         self.gating_strategy.add_gate(gate)
 
+    def add_sample(self, samples):
+        self.samples.extend(load_samples(samples))
+
     def add_transform(self, transform):
         self.gating_strategy.add_transform(transform)
 
@@ -239,6 +246,7 @@ def analyze_samples(self, verbose=False):
         # indices for each sample. Add convenience functions inside
         # GatingResults class for conveniently extracting information from
         # the DataFrame
+        # TODO: should this method take a subsample option to gate on just the sub-sampled events?
         results = gate_samples(self.gating_strategy, self.samples, verbose)
 
         all_reports = [res.report for res in results]

setup.py

@@ -16,7 +16,7 @@
 
 setup(
     name='FlowKit',
-    version='0.3.1',
+    version='0.3.2',
     packages=find_packages(),
     package_data={'': ['*.xsd']},
     include_package_data=True,

unit_test/sample_unit_tests.py

@@ -90,6 +90,27 @@ def test_transform_sample_hyperlog(self):
 
         self.assertIsInstance(data1_sample._transformed_events, np.ndarray)
 
+    def test_create_fcs(self):
+        fcs_file_path = "examples/test_comp_example.fcs"
+        comp_file_path = Path("examples/comp_complete_example.csv")
+
+        sample = Sample(
+            fcs_path_or_data=fcs_file_path,
+            compensation=comp_file_path
+        )
+
+        sample.export("test_fcs_export.fcs", source='comp', directory="examples")
+
+        exported_fcs_file = "examples/test_fcs_export.fcs"
+        exported_sample = Sample(fcs_path_or_data=exported_fcs_file)
+        os.unlink(exported_fcs_file)
+
+        self.assertIsInstance(exported_sample, Sample)
+
+        # TODO: Excluding time channel here, as the difference was nearly 0.01. Need to investigate why the
+        #       exported comp data isn't exactly equal
+        np.testing.assert_almost_equal(sample._comp_events[:, :-1], exported_sample._raw_events[:, :-1], decimal=3)
+
 
 if __name__ == '__main__':
     unittest.main()