dclab 0.62.12__tar.gz → 0.62.14__tar.gz

{dclab-0.62.12 → dclab-0.62.14}/CHANGELOG

@@ -1,3 +1,8 @@
+0.62.14
+ - fix: exporting table data broken
+ - ref: introduce base class for tables `TableMeta`
+0.62.13
+ - ref: add custom table class for `hdf5` and `dcor` formats
 0.62.12
  - fix: viscosity computation for MP-PBS at zero flow rate should be NaN
 0.62.11

{dclab-0.62.12 → dclab-0.62.14}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: dclab
-Version: 0.62.12
+Version: 0.62.14
 Summary: Library for real-time deformability cytometry (RT-DC)
 Author: Benedikt Hartmann, Eoghan O'Connell, Maik Herbig, Maximilian Schlögel, Nadia Sbaa, Paul Müller, Philipp Rosendahl, Raghava Alajangi
 Maintainer-email: Paul Müller <dev@craban.de>

{dclab-0.62.12 → dclab-0.62.14}/dclab/_version.py

@@ -12,5 +12,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.62.12'
-__version_tuple__ = version_tuple = (0, 62, 12)
+__version__ = version = '0.62.14'
+__version_tuple__ = version_tuple = (0, 62, 14)

dclab-0.62.14/dclab/rtdc_dataset/fmt_dcor/tables.py

@@ -0,0 +1,66 @@
+import numpy as np
+
+from ..meta_table import MetaTable
+
+
+class DCORTables:
+    def __init__(self, api):
+        self.api = api
+        self._tables_cache = None
+
+    def __contains__(self, key):
+        return key in self.keys()
+
+    def __getitem__(self, key):
+        return self._tables[key]
+
+    def __iter__(self):
+        for key in self.keys():
+            yield key
+
+    def __len__(self):
+        return len(self._tables)
+
+    def keys(self):
+        return self._tables.keys()
+
+    @property
+    def _tables(self):
+        if self._tables_cache is None:
+            table_data = self.api.get(query="tables")
+            # assemble the tables
+            tables = {}
+            for key in table_data:
+                tables[key] = DCORTable(table_data[key])
+
+            self._tables_cache = tables
+        return self._tables_cache
+
+
+class DCORTable(MetaTable):
+    def __init__(self, table_content):
+        self._columns, data = table_content
+        self._tab_data = np.asarray(data)
+        if self._columns is not None:
+            # We have a rec-array (named columns)
+
+            ds_dt = np.dtype({'names': self._columns,
+                              'formats': [np.float64] * len(self._columns)})
+            self._tab_data = np.rec.array(self._tab_data, dtype=ds_dt)
+
+    def __array__(self, *args, **kwargs):
+        return self._tab_data.__array__(*args, **kwargs)
+
+    @property
+    def meta(self):
+        # TODO: Implement metadata sending from DCOR.
+        return {}
+
+    def has_graphs(self):
+        return self._columns is not None
+
+    def keys(self):
+        return self._columns
+
+    def __getitem__(self, key):
+        return self._tab_data[key]

{dclab-0.62.12 → dclab-0.62.14}/dclab/rtdc_dataset/fmt_hdf5/tables.py

@@ -1,3 +1,6 @@
+from ..meta_table import MetaTable
+
+
 class H5Tables:
     def __init__(self, h5):
         self.h5file = h5
@@ -5,7 +8,7 @@ class H5Tables:
 
     def __getitem__(self, key):
         if key in self.keys():
-            tab = self.h5file["tables"][key]
+            tab = H5Table(self.h5file["tables"][key])
         else:
             raise KeyError(f"Table '{key}' not found or empty "
                            f"in {self.h5file.file.filename}!")
@@ -28,3 +31,30 @@ class H5Tables:
                         names.append(key)
             self._cache_keys = names
         return self._cache_keys
+
+
+class H5Table(MetaTable):
+    def __init__(self, h5_ds):
+        self._h5_ds = h5_ds
+        self._keys = None
+        self._meta = None
+
+    def __array__(self, *args, **kwargs):
+        return self._h5_ds.__array__(*args, **kwargs)
+
+    @property
+    def meta(self):
+        if self._meta is None:
+            self._meta = dict(self._h5_ds.attrs)
+        return self._meta
+
+    def has_graphs(self):
+        return self.keys() is not None
+
+    def keys(self):
+        if self._keys is None:
+            self._keys = self._h5_ds.dtype.names
+        return self._keys
+
+    def __getitem__(self, key):
+        return self._h5_ds[key]

dclab-0.62.14/dclab/rtdc_dataset/meta_table.py

@@ -0,0 +1,25 @@
+import abc
+
+
+class MetaTable(abc.ABC):
+    @abc.abstractmethod
+    def __array__(self, *args, **kwargs):
+        """Return array representation of the table"""
+
+    @property
+    @abc.abstractmethod
+    def meta(self):
+        """Return metadata of the table (e.g. graph colors)"""
+
+    @abc.abstractmethod
+    def has_graphs(self):
+        """Return True when the table has key-based graphs"""
+
+    @abc.abstractmethod
+    def keys(self):
+        """Return keys of the graphs, None if `not self.has_graphs()`"""
+
+    @abc.abstractmethod
+    def __getitem__(self, key):
+        """Return a graph or otherwise part of the data array"""
+        pass

{dclab-0.62.12 → dclab-0.62.14}/dclab/rtdc_dataset/writer.py

@@ -17,6 +17,7 @@ from ..util import hashobj
 from .._version import version
 
 from .feat_anc_plugin import PlugInFeature
+from .meta_table import MetaTable
 
 #: DEPRECATED (use `CHUNK_SIZE_BYTES` instead)
 CHUNK_SIZE = 100
@@ -641,6 +642,12 @@ class RTDCWriter:
         if h5_attrs is None:
             h5_attrs = {}
 
+        # Convert MetaTable to numpy data
+        if isinstance(cmp_array, MetaTable):
+            h5_attrs.update(cmp_array.meta)
+            cmp_array = cmp_array.__array__()
+
+        # Handle individual cases
         if isinstance(cmp_array, np.recarray):
             # A table is a compound array (np.recarray). If we are here,
             # this means that the user passed an instance of np.recarray.

{dclab-0.62.12 → dclab-0.62.14}/dclab.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: dclab
-Version: 0.62.12
+Version: 0.62.14
 Summary: Library for real-time deformability cytometry (RT-DC)
 Author: Benedikt Hartmann, Eoghan O'Connell, Maik Herbig, Maximilian Schlögel, Nadia Sbaa, Paul Müller, Philipp Rosendahl, Raghava Alajangi
 Maintainer-email: Paul Müller <dev@craban.de>

{dclab-0.62.12 → dclab-0.62.14}/dclab.egg-info/SOURCES.txt

@@ -102,6 +102,7 @@ dclab/rtdc_dataset/fmt_dict.py
 dclab/rtdc_dataset/fmt_http.py
 dclab/rtdc_dataset/fmt_s3.py
 dclab/rtdc_dataset/load.py
+dclab/rtdc_dataset/meta_table.py
 dclab/rtdc_dataset/writer.py
 dclab/rtdc_dataset/feat_anc_core/__init__.py
 dclab/rtdc_dataset/feat_anc_core/af_basic.py

{dclab-0.62.12 → dclab-0.62.14}/tests/test_rtdc_export_hdf5.py

@@ -582,7 +582,7 @@ def test_hdf5_override():
 def test_hdf5_tables(tables):
     keys = ["area_um", "deform", "time", "frame", "index_online"]
     ddict = example_data_dict(size=10, keys=keys)
-    ds1 = dclab.new_dataset(ddict)
+    ds1 = new_dataset(ddict)
     ds1.config["experiment"]["sample"] = "test"
     ds1.config["experiment"]["run index"] = 1
     ds1.config["imaging"]["frame rate"] = 2000
@@ -608,11 +608,18 @@ def test_hdf5_tables(tables):
         else:
             assert "src_iratrax" not in h5.get("tables", {})
 
+    if tables:
+        # reading the file with dclab
+        with new_dataset(f1) as ds:
+            assert "src_iratrax" in ds.tables
+            assert ds.tables["src_iratrax"].has_graphs()
+            assert np.all(ds.tables["src_iratrax"] == rec_arr)
+
 
 def test_hdf5_tables_array_only():
     keys = ["area_um", "deform", "time", "frame", "index_online"]
     ddict = example_data_dict(size=10, keys=keys)
-    ds1 = dclab.new_dataset(ddict)
+    ds1 = new_dataset(ddict)
     ds1.config["experiment"]["sample"] = "test"
     ds1.config["experiment"]["run index"] = 1
     ds1.config["imaging"]["frame rate"] = 2000
@@ -629,6 +636,63 @@ def test_hdf5_tables_array_only():
         assert np.all(h5["tables"]["src_iratrax"] == tab_data)
         assert h5["tables"]["src_iratrax"].attrs["CLASS"] == np.bytes_("IMAGE")
 
+    # reading the file with dclab
+    with new_dataset(f1) as ds:
+        assert "src_iratrax" in ds.tables
+        assert not ds.tables["src_iratrax"].has_graphs()
+        assert np.all(ds.tables["src_iratrax"] == tab_data)
+        assert ds.tables["src_iratrax"].meta["CLASS"] == np.bytes_("IMAGE")
+
+
+def test_hdf5_table_from_file(tmp_path):
+    ds = new_dataset(retrieve_data("fmt-hdf5_raw-cytoshot-exported.zip"))
+
+    f1 = join(tmp_path, "test.rtdc")
+    ds.export.hdf5(f1, ["temp"], tables=True)
+
+    ds2 = new_dataset(f1)
+    tab = ds2.tables["src_so2exp_cytoshot_monitor"]
+
+    assert np.allclose(
+        tab["focus"][0],
+        6.241900073522602,
+        atol=1e-7, rtol=0)
+
+    assert tab.meta["COLOR_focus"] == "#a03000"
+
+
+def test_hdf5_table_from_file_with_unnamed_table(tmp_path):
+    path = retrieve_data("fmt-hdf5_raw-cytoshot-exported.zip")
+    with RTDCWriter(path) as hw:
+        hw.store_table("just_an_array", np.ones((30, 50)),
+                       h5_attrs={"heinz": "kunz"})
+
+    # Make sure that worked with tables
+    with new_dataset(path) as ds:
+        tab = ds.tables["just_an_array"]
+        assert not tab.has_graphs()
+        assert np.array(tab).shape == (30, 50)
+        assert tab.meta["heinz"] == "kunz"
+
+    with new_dataset(path) as ds:
+        f1 = join(tmp_path, "test.rtdc")
+        ds.export.hdf5(f1, ["temp"], tables=True)
+
+    with new_dataset(f1) as ds2:
+        tab = ds2.tables["src_so2exp_cytoshot_monitor"]
+        assert tab.has_graphs()
+        assert np.allclose(
+            tab["focus"][0],
+            6.241900073522602,
+            atol=1e-7, rtol=0)
+
+        assert tab.meta["COLOR_focus"] == "#a03000"
+
+        tab2 = ds2.tables["src_just_an_array"]
+        assert not tab2.has_graphs()
+        assert np.array(tab2).shape == (30, 50)
+        assert tab2.meta["heinz"] == "kunz"
+
 
 def test_hdf5_trace_from_tdms():
     pytest.importorskip("nptdms")

dclab-0.62.12/dclab/rtdc_dataset/fmt_dcor/tables.py

@@ -1,42 +0,0 @@
-import numpy as np
-
-
-class DCORTables:
-    def __init__(self, api):
-        self.api = api
-        self._tables_cache = None
-
-    def __contains__(self, key):
-        return key in self.keys()
-
-    def __getitem__(self, key):
-        return self._tables[key]
-
-    def __iter__(self):
-        for key in self.keys():
-            yield key
-
-    def __len__(self):
-        return len(self._tables)
-
-    def keys(self):
-        return self._tables.keys()
-
-    @property
-    def _tables(self):
-        if self._tables_cache is None:
-            table_data = self.api.get(query="tables")
-            # assemble the tables
-            tables = {}
-            for key in table_data:
-                columns, data = table_data[key]
-                tab_data = np.asarray(data)
-                if columns is not None:
-                    # We have a rec-array (named columns)
-                    ds_dt = np.dtype({'names': columns,
-                                      'formats': [np.float64] * len(columns)})
-                    tab_data = np.rec.array(tab_data, dtype=ds_dt)
-                tables[key] = tab_data
-
-            self._tables_cache = tables
-        return self._tables_cache