rosetta-eclab 0.1.3__tar.gz → 0.1.4__tar.gz

{rosetta_eclab-0.1.3 → rosetta_eclab-0.1.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rosetta-eclab
-Version: 0.1.3
+Version: 0.1.4
 Summary: Name the columns of BioLogic EC-Lab .mpr files that your parser does not know - no language model, evidence for every answer
 Author: Nataniel Ataseven
 License-Expression: GPL-3.0-or-later
@@ -139,7 +139,7 @@ The namer improved between hold-outs, so the total includes earlier versions. Th
 - **Values EC-Lab computes after the fact** cannot be reconstructed reliably. For example, a capacitance is held per segment.
 - **Noise-like DC currents** in impedance records, and **external analog inputs** with no sibling column, usually get ABSTAIN.
 - **Two quantities with the same definition in a file** cannot be told apart from the data. In galvanostatic cycling, for example, `dQ` and `Q charge or discharge` reset at the same points. A prior over how common each name is for the technique decides between them, and the alternative is shown.
-- **Supported formats:** EC-Lab data-module versions 2, 3, 10 and 11. Very old files (version 0) are not supported yet.
+- **Supported formats:** EC-Lab data-module versions 0, 2, 3, 10 and 11. Version 0 (older EC-Lab) is read with galvani's offsets, because yadg does not read it. All 59 public version-0 files decode byte for byte like galvani 0.5.0, and the two with an EC-Lab text export match it. The file technique is not read from version-0 files, so the namer works without it.
 
 ## Reproduce the benchmark
 

{rosetta_eclab-0.1.3 → rosetta_eclab-0.1.4}/README.md

@@ -114,7 +114,7 @@ The namer improved between hold-outs, so the total includes earlier versions. Th
 - **Values EC-Lab computes after the fact** cannot be reconstructed reliably. For example, a capacitance is held per segment.
 - **Noise-like DC currents** in impedance records, and **external analog inputs** with no sibling column, usually get ABSTAIN.
 - **Two quantities with the same definition in a file** cannot be told apart from the data. In galvanostatic cycling, for example, `dQ` and `Q charge or discharge` reset at the same points. A prior over how common each name is for the technique decides between them, and the alternative is shown.
-- **Supported formats:** EC-Lab data-module versions 2, 3, 10 and 11. Very old files (version 0) are not supported yet.
+- **Supported formats:** EC-Lab data-module versions 0, 2, 3, 10 and 11. Version 0 (older EC-Lab) is read with galvani's offsets, because yadg does not read it. All 59 public version-0 files decode byte for byte like galvani 0.5.0, and the two with an EC-Lab text export match it. The file technique is not read from version-0 files, so the namer works without it.
 
 ## Reproduce the benchmark
 

{rosetta_eclab-0.1.3 → rosetta_eclab-0.1.4}/namer/library.py

@@ -561,7 +561,11 @@ def build() -> dict[str, Quantity]:
     q("z cycle", shapes=[Shape("integer counter in a sequence with impedance", lambda c: 0.75 * is_counter(c) if c.col("freq") is not None else 0.0)])
     q("Ns", shapes=[Shape("sequence index from 0 that changes with the control", lambda c: ns_shape(c))])
     q("I Range", shapes=[Shape("instrument range code (few values, not starting at 0)", lambda c: irange_shape(c))])
-    q("Pwe", relations=[Relation("Ewe*I", ("Ewe", CURRENT), lambda c: _e(c) * _i(c), [W_PER_V_MA])])
+    q("Pwe", relations=[
+        Relation("Ewe*I", ("Ewe", CURRENT), lambda c: _e(c) * _i(c), [W_PER_V_MA]),
+        # some older EC-Lab files store the power's magnitude (galvani's bio_logic1.mpr and bio_logic3.mpr)
+        Relation("|Ewe*I|", ("Ewe", CURRENT), lambda c: np.abs(_e(c) * _i(c)), [W_PER_V_MA]),
+    ])
     q("Ewe", relations=[
         Relation("Pwe/I", ("Pwe", ("I", "dq", "(Q-Qo)")), lambda c: c.col("Pwe") / _i(c), [1000.0]),
         Relation("follows potential control", ("control",), lambda c: c.col("control"), [1.0]),

{rosetta_eclab-0.1.3 → rosetta_eclab-0.1.4}/namer/mprfile.py

@@ -131,10 +131,19 @@ def read(path: str | Path, reference: str = "galvani") -> MprData:
     elif version in (2, 3):
         yids = full = np.frombuffer(data, offset=5, dtype="<u2", count=n_columns).tolist()
         offset = 0x195 if version == 2 else 0x196
+    elif version == 0:  # older EC-Lab files; offsets as in galvani (yadg 7.0.1 does not read version 0)
+        if data[5]:  # one byte per column ID; all public version-0 files use this layout
+            yids = full = list(data[5:5 + n_columns])
+            offset = 100
+        else:  # galvani: EC-Lab 11.50+ writes [0 1 0 3 0 174 ...]; no public file to test against
+            yids = full = list(data[6:5 + 2 * n_columns:2])
+            offset = 1007
     else:
         raise NotImplementedError(f"data module version {version} is not supported yet")
+    if n_points == 0:
+        raise ValueError("the file has no data points")
     region = len(data) - offset
-    if n_points == 0 or region % n_points:
+    if region % n_points:
         raise ValueError(f"record size is not a whole number of bytes ({region} / {n_points})")
     rec = region // n_points
 
@@ -142,6 +151,9 @@ def read(path: str | Path, reference: str = "galvani") -> MprData:
     gref = galvani_ref.reference() if reference == "galvani" else None
     gmap = gref.colmap if gref else {}
     notes = _galvani_notes(full, gref, version in (10, 11)) if gref else []
+    if version == 0 and reference == "yadg":
+        notes.append("yadg 7.0.1 cannot read this file at all (data module version 0); "
+                     "'known' here means listed in yadg's column map")
     layout = []  # record layout in file order: (raw_id or None for the flags byte, ...)
     for yid, fid in zip(yids, full):
         if yid in FLAG_IDS:  # all flag IDs share one byte, at the position of the first one

{rosetta_eclab-0.1.3 → rosetta_eclab-0.1.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "rosetta-eclab"
-version = "0.1.3"
+version = "0.1.4"
 description = "Name the columns of BioLogic EC-Lab .mpr files that your parser does not know - no language model, evidence for every answer"
 readme = "README.md"
 license = "GPL-3.0-or-later"

{rosetta_eclab-0.1.3 → rosetta_eclab-0.1.4}/rosetta_eclab.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rosetta-eclab
-Version: 0.1.3
+Version: 0.1.4
 Summary: Name the columns of BioLogic EC-Lab .mpr files that your parser does not know - no language model, evidence for every answer
 Author: Nataniel Ataseven
 License-Expression: GPL-3.0-or-later
@@ -139,7 +139,7 @@ The namer improved between hold-outs, so the total includes earlier versions. Th
 - **Values EC-Lab computes after the fact** cannot be reconstructed reliably. For example, a capacitance is held per segment.
 - **Noise-like DC currents** in impedance records, and **external analog inputs** with no sibling column, usually get ABSTAIN.
 - **Two quantities with the same definition in a file** cannot be told apart from the data. In galvanostatic cycling, for example, `dQ` and `Q charge or discharge` reset at the same points. A prior over how common each name is for the technique decides between them, and the alternative is shown.
-- **Supported formats:** EC-Lab data-module versions 2, 3, 10 and 11. Very old files (version 0) are not supported yet.
+- **Supported formats:** EC-Lab data-module versions 0, 2, 3, 10 and 11. Version 0 (older EC-Lab) is read with galvani's offsets, because yadg does not read it. All 59 public version-0 files decode byte for byte like galvani 0.5.0, and the two with an EC-Lab text export match it. The file technique is not read from version-0 files, so the namer works without it.
 
 ## Reproduce the benchmark
 

{rosetta_eclab-0.1.3 → rosetta_eclab-0.1.4}/tests/test_cli.py

@@ -45,6 +45,46 @@ def test_reader_matches_reference_decoder():
         assert a == b, name
 
 
+def _mpt(path: Path) -> dict[str, np.ndarray]:
+    lines = path.read_bytes().decode("latin-1").splitlines()
+    n_header = int(lines[1].split(":")[1])
+    head = lines[n_header - 1].rstrip("\t").split("\t")
+    rows = [line.rstrip("\t").split("\t") for line in lines[n_header:] if line.strip()]
+    return {h: np.array([float(r[i].replace(",", ".")) for r in rows]) for i, h in enumerate(head)}
+
+
+def test_version_0_files_match_the_vendor_export():
+    """Data module version 0 (older EC-Lab): yadg cannot read it; the values must match EC-Lab's own .mpt."""
+    m = mprfile.read(EXT / "bio_logic1.mpr", reference="galvani")
+    mpt = _mpt(EXT / "bio_logic1.mpt")
+    cols = {c.name: c.values.astype("f8") for c in m.columns}
+    assert m.n_points == len(mpt["time/s"]) == 3119 and not m.unknown
+    for name, ref in (("time", "time/s"), ("Ewe", "Ewe/V"), ("dq", "dq/mA.h"), ("Pwe", "P/W")):
+        np.testing.assert_allclose(cols[name], mpt[ref], rtol=1e-6, atol=1e-9, err_msg=name)
+    assert any("yadg 7.0.1 cannot read" in n for n in mprfile.read(EXT / "bio_logic1.mpr", reference="yadg").notes)
+    assert not any("yadg" in n for n in m.notes)
+
+
+def test_empty_file_is_reported_plainly():
+    with pytest.raises(ValueError, match="no data points"):
+        mprfile.read(EXT / "wait.mpr")
+
+
+def test_power_magnitude_in_version_0_file_is_named(tmp_path):
+    """Version-0 files store |Ewe*I| as the power; hidden, it must come back as Pwe (it was named Ece)."""
+    raw = bytearray((EXT / "bio_logic1.mpr").read_bytes())
+    mdata = next(md for name, _v, md in mprfile._modules(bytes(raw)) if name == "VMP data")
+    start = bytes(raw).index(mdata)
+    ids = list(raw[start + 5: start + 5 + raw[start + 4]])
+    raw[start + 5 + ids.index(70)] = 255  # one-byte column IDs in version 0
+    f = tmp_path / "old_drift.mpr"
+    f.write_bytes(bytes(raw))
+    m = mprfile.read(f, reference="galvani")
+    assert [c.raw_id for c in m.unknown] == [255]
+    res = cli.name_unknown(m)
+    assert res[0]["verdict"] == "NAME" and res[0]["name"] == "Pwe", res[0]
+
+
 def test_real_drift_duplicates_are_reported():
     m = mprfile.read(EXT / "gcpl.issue_230.mpr", reference="galvani")
     assert [c.raw_id for c in m.unknown] == [883, 884]