eclab.mpr: Add column conflict resolution scheme. (#227)

* Fix columns

* Add new tests.

* conflict_columns

* docs changes.

* typo

Author: PeterKraus

docs/source/version.6_2.rst

@@ -23,3 +23,4 @@ Bug fixes in ``yadg-next`` include:
   - Added further ``Set I/C`` parameters. Thanks to `@Locki3 <https://github.com/Locki3>`_ for providing test files.
   - Fixed ``param format`` and ``data_column`` in CV files generated with EC-Lab version 11.50 using the :mod:`yadg.extractors.eclab.mpr` module. Thank you to J.N. Hausmann from Helmholtz-Zentrum Berlin für Materialien und Energie for providing the test files.
   - Fixed parsing of various kinds of Modulo Bat files in :mod:`yadg.extractors.eclab.mpr`. Thank you to `@JohannesBaller <https://github.com/JohannesBaller>`_ for providing test files.
+  - Columns in :mod:`yadg.extractors.eclab.mpr` files may have different meanings based on which other columns are also present in the files, see `issue 225 <https://github.com/dgbowl/yadg/issues/225>`_. Added a (hopefully extensible) way to tackle such conflicts and clarified the warnings.

src/yadg/extractors/eclab/mpr.py

@@ -199,6 +199,7 @@
     settings_dtypes,
     flag_columns,
     data_columns,
+    conflict_columns,
     log_dtypes,
     extdev_dtypes,
 )
@@ -314,17 +315,57 @@ def parse_columns(column_ids: list[int]) -> tuple[list, list, list, dict]:
                 names.append("flags")
                 dtypes.append("|u1")
                 units.append(None)
-        elif id in data_columns:
+        elif id in data_columns and id not in conflict_columns:
             dtype, name, unit = data_columns[id]
             if name in names:
-                logger.warning("Duplicate column '%s' with unit '%s'.", name, unit)
+                logger.error(
+                    "Column ID %d is a duplicate of '%s' with unit '%s'. "
+                    "This is most certainly an error, please submit a bug report.",
+                    id,
+                    name,
+                    unit,
+                )
                 name = f"duplicate {name}"
             names.append(name)
             dtypes.append(dtype)
             units.append(unit)
+        elif id in conflict_columns:
+            for cid, cvals in conflict_columns[id].items():
+                if cid in column_ids:
+                    dtype, name, unit = cvals
+                    names.append(name)
+                    dtypes.append(dtype)
+                    units.append(unit)
+                    logger.warning(
+                        "Ambiguous column ID %d assigned as '%s' with unit '%s'. "
+                        "Please check this assignment manually.",
+                        id,
+                        name,
+                        unit,
+                    )
+                    break
+            else:
+                dtype, name, unit = data_columns[id]
+                if name in names:
+                    logger.error(
+                        "Column ID %d is duplicate of '%s' with unit '%s'. "
+                        "This is most certainly an error, please submit a bug report.",
+                        id,
+                        name,
+                        unit,
+                    )
+                    name = f"duplicate {name}"
+                names.append(name)
+                dtypes.append(dtype)
+                units.append(unit)
         else:
             name = f"unknown_{len(names)}"
-            logger.warning("Unknown column ID %d was assigned into '%s'.", id, name)
+            logger.error(
+                "Unknown column ID %d encountered and assigned into '%s'. "
+                "This is an error, please submit a bug report.",
+                id,
+                name,
+            )
             names.append(name)
             dtypes.append("<f4")
             units.append(None)

src/yadg/extractors/eclab/mpr_columns.py

@@ -110,12 +110,13 @@
     169: ("<f4", "Cs", "µF"),
     172: ("<f4", "Cp", "µF"),
     173: ("<f4", "Cp⁻²", "µF⁻²"),
-    174: ("<f4", "<Ewe>", "V"),
+    174: ("<f4", "<Ewe>", "V"),  # This column may conflict with ID 77.
     178: ("<f4", "(Q-Qo)", "C"),
     179: ("<f4", "dQ", "C"),
     182: ("<f8", "step time", "s"),
     185: ("<f4", "<Ece>", "V"),
     211: ("<f8", "Q charge or discharge", "C"),
+    215: ("<f4", "<Ece>", "V"),
     217: ("<f4", "THD Ewe", "%"),
     241: ("<f4", "|E1|", "V"),
     242: ("<f4", "|E2|", "V"),
@@ -128,6 +129,7 @@
     326: ("<f4", "P", "W"),
     331: ("<f4", "Re(Z1)", "Ω"),
     332: ("<f4", "Re(Z2)", "Ω"),
+    352: ("<f4", "|Ece|", "Ω"),
     361: ("<f4", "-Im(Z1)", "Ω"),
     362: ("<f4", "-Im(Z2)", "Ω"),
     368: ("<f8", "Energy we", "W·h"),
@@ -181,6 +183,13 @@
     981: ("<u4", "z cycle", None),  # 981 % 512 = 469, also z cycle
 }
 
+# Conflict resolution map. If both the outer and inner column ID are present,
+# the meaning of the outer column ID is set to the entry below.
+conflict_columns = {
+    174: {
+        77: ("<f4", "Phase(Zwe-ce)", "deg"),
+    },
+}
 
 # Relates the offset in log data to the corresponding dtype and name.
 # NOTE: The safety limits are maybe at 0x200?

src/yadg/extractors/eclab/mpt.py

@@ -245,6 +245,8 @@ def process_data(
     units = dict()
     columns = list()
     for n in names:
+        if n.strip() == "":
+            continue
         c, u = column_units[n.strip()]
         if c in columns:
             logger.warning("Duplicate column '%s' with unit '%s'.", c, u)

src/yadg/extractors/eclab/mpt_columns.py

@@ -13,6 +13,7 @@
     "<Ece>/V": ("<Ece>", "V"),
     "<Ewe>/V": ("<Ewe>", "V"),
     "<Ewe/V>": ("<Ewe>", "V"),
+    "<Ewe-Ece>/V": ("<Ewe-Ece>", "V"),
     "<I>/mA": ("<I>", "mA"),
     "|C|/nF": ("|C|", "nF"),
     "|Conductivity|/mS/cm": ("|Conductivity|", "mS/cm"),

src/yadg/extractors/eclab/techniques.py

@@ -1697,7 +1697,7 @@ def dev_derived(
     elif unit in {"deg"}:
         # VMP-3: using accuracy: 1 degree
         return 1.0
-    elif unit in {"Ω", "S", "W"}:
+    elif unit in {"Ω", "S", "W", "Ω⁻¹"}:
         # [Ω] = [V]/[A];
         # [S] = [A]/[V];
         # [W] = [A]*[V];

tests/test_x_eclab.py

@@ -69,6 +69,9 @@ def compare_params(left, right):
         ("ocv", "en_US"),
         ("ocv.issue_149", "de_DE"),
         ("peis", "en_US"),
+        ("peis.issue_225.no_ece", "en_US"),
+        ("peis.issue_225.with_ece", "en_US"),
+        ("peis.issue_225.ewe_ece", "en_US"),
         ("wait", "en_US"),
         ("vsp_ocv_wo", "en_US"),
         ("vsp_ocv_with", "en_US"),
@@ -88,7 +91,7 @@ def test_eclab_consistency(froot, locale, datadir):
         try:
             xr.testing.assert_allclose(aret[key], bret[key])
         except AssertionError as e:
-            e.args = (e.args[0] + f"Error happened on key: {key!r}\n",)
+            e.args = (e.args[0] + f"\nError happened on key: {key!r}\n",)
             raise e
     compare_params(aret, bret)
 
@@ -111,7 +114,7 @@ def test_eclab_consistency_extract_mpr_bytes(froot, locale, datadir):
         try:
             xr.testing.assert_equal(aret[key], bret[key])
         except AssertionError as e:
-            e.args = (e.args[0] + f"Error happened on key: {key!r}\n",)
+            e.args = (e.args[0] + f"\nError happened on key: {key!r}\n",)
             raise e
     compare_params(aret, bret)
 
@@ -149,7 +152,7 @@ def test_eclab_consistency_partial_1(froot, locale, datadir):
         try:
             xr.testing.assert_allclose(aret[key], bret[key])
         except AssertionError as e:
-            e.args = (e.args[0] + f"Error happened on key: {key!r}\n",)
+            e.args = (e.args[0] + f"\nError happened on key: {key!r}\n",)
             raise e
     compare_params(aret, bret)
 
@@ -176,7 +179,7 @@ def test_eclab_consistency_partial_2(froot, locale, datadir):
         try:
             xr.testing.assert_allclose(aret[key], bret[key])
         except AssertionError as e:
-            e.args = (e.args[0] + f"Error happened on key: {key!r}\n",)
+            e.args = (e.args[0] + f"\nError happened on key: {key!r}\n",)
             raise e
     compare_params(aret, bret)
 
@@ -217,5 +220,5 @@ def test_eclab_mpt_old(afile, bfile, datadir):
             if key in {"Efficiency", "Efficiency_std_err"}:
                 pass
             else:
-                e.args = (e.args[0] + f"Error happened on key: {key!r}\n",)
+                e.args = (e.args[0] + f"\nError happened on key: {key!r}\n",)
                 raise e