Fix Various Warnings (#460)

* fix deprecated datetime.datetime.utcnow call * fix deprecated datetime.datetime.utcnow call * fix deprecated datetime.datetime.utcnow call * fix deprecated datetime.datetime.utcnow call * fix deprecated numpy.in1d call * catch and filter RankWarning and OptimizeWarning * only include df for concat if not empty, avoids pandas FutureWarning * import snuck in there * just format here * this should work around the unavailability of Rankwarning on 3.9 * also catch out the OptimizeWarning - REVERT if not desired * imports here * context manager to ignore the NaturalNameWarning - RESET if we wish to keep them * a tiny bit of formatting * FIX Series.__getitem__ warning by using recommended .iloc * Introduce CI extras just for our GA workflows, to speed up without polluting developper requirements * install with CI extra and set auto workers determination * some type checking * catch and ignore UserWarning: The figure layout has changed to tight because we do it on purpose * verbose flag to get some more output * import order * FIX UserWarning: set_ticklabels() etc by setting both the ticks and labels at the same time, as recommended * FIX UserWarning: : No artists with labels found to put in legend by catching, this can happen due to leading underscores in labels * remove unused imports * cleaner * try using only 2 to spare the tests triggering multiprocessing * unused imports * changelog entries * patch version * typo * let's forget about this for now * have to remove this too * and this part too * bump to 0.16.1 and target a merge after #458 * make linting happy * remove this too * ask coverage workflow to display the coverage stats * add citation instruction to doc index * do not bare catch * let's let linters complain * expand for josch * better as list comprehension * also raise for the user * filter out warnings and log them * filter out warnings and log them * do not intercept this * filter out warnings and log them * unused imports * more cleanup * specify new behaviour in changelog * oops typo * remove this comment * should be today * dotdict
pylhc · Sep 20, 2024 · 95024bc · 95024bc
1 parent d8c8923
commit 95024bc
Show file tree

Hide file tree

Showing 19 changed files with 306 additions and 169 deletions.
diff --git a/.github/workflows/coverage.yml b/.github/workflows/coverage.yml
@@ -12,5 +12,5 @@ jobs:
       uses: pylhc/.github/.github/workflows/coverage.yml@master
       with:
         src-dir: omc3
-        pytest-options: -m "not cern_network"
+        pytest-options: -m "not cern_network" --cov-report term-missing
       secrets: inherit
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,5 +1,21 @@
 # OMC3 Changelog
 
+#### 2024-09-20 - v0.16.1 - _fsoubelet_
+
+- Fixed:
+  - Fixed `DepracationWarning`s related datetime operations.
+  - Fixed `DeprecationWarning` occuring due to the use of old `numpy` functions.
+  - Fixed `FutureWarning` happening during edge-cases of dataframe concatenation by performing checks ahead of time.
+  - Fixed `FutureWarning`s occuring due to deprecated `pandas.Series` accesses.
+  - Fixed `UserWarning` occuring when wrongly setting ticks and labels for correction plots.
+
+- Changed:
+  - Masked `NaturalNameWarning`s happening during HDF5 tables operations, as the use of names such as `kq4.r8b2` is not avoidable and `pandas` properly handles access operations for us.
+  - Masked `UserWarning`s happening during plotting for operations that are explicitely demanded.
+  - Intercept `RankWarning` which can happen during a `polyfit` of data and re-emit as log message.
+  - Intercept `OptimizeWarning` happening when the covariance parameters could not be estimated in `kmod` analysis and re-emit as log message.
+  - Intercept `OptimizeWarning` happening when the covariance parameters could not be estimated in `rdt` analysis and re-emit as log message.
+
 #### 2024-09-18 - v0.16.0 - _jdilly_
 
 - Added:
@@ -94,7 +110,7 @@
 #### 2023-11-29 - v0.12.0 - _jdilly_
 
 - Added to harmonic analysis:
-  - `suffix` input parameter: adds suffix to output files, which e.g. allows running the same file 
+  - `suffix` input parameter: adds suffix to output files, which e.g. allows running the same file
     with different parameters without overwriting it.
   - `bunch_ids` input parameter: in case of multibunch-files only analyse these bunches.
     If not given, all bunches will be analysed, as before.
@@ -117,15 +133,15 @@
   - Plot Optics: making normalized dispersion plot a special case.
 
 - Added:
-  - Plot Optics: optional input "--labels" to manually set the legend-labels. 
+  - Plot Optics: optional input "--labels" to manually set the legend-labels.
 
 #### 2023-06-16 - v0.11.1 - _jdilly_
 
 - Fixed:
   - OptionalString: 'None' as input is converted to None.
   - Missing Kerberos config added to MANIFEST for packaging.
   - Plot Optics plots now correct error-column, e.g. for beta-beating.
-  - Added warnings/errors for too few bpms in N-BPM/3-BPM methods. 
+  - Added warnings/errors for too few bpms in N-BPM/3-BPM methods.
   - Added navbar to sphinx documentation.
 
 - Tests:
@@ -170,12 +186,11 @@
 #### 2023-01-20 - v0.7.1 - _jdilly_
 
 - Added:
-  - Amplitude Detuning plots: Switch to plot only with/without BBQ correction 
+  - Amplitude Detuning plots: Switch to plot only with/without BBQ correction.
 
 - Fixed:
   - Second Order Amplitude Detuning fit now working
-  - Correct print/calculation of second order direct terms for forced 
-    kicks in plot-labels.
+  - Correct print/calculation of second order direct terms for forced kicks in plot-labels.
 
 #### 2022-11-08 - v0.7.0 - _jdilly_
 
@@ -187,14 +202,14 @@
 
 #### 2022-11-01 - v0.6.6
 
-- Bugfixes 
+- Bugfixes:
   - correction: fullresponse is converted to Path.
-  - fake measurement from model: dont randomize errors and values by default. 
+  - fake measurement from model: dont randomize errors and values by default.
 
 #### 2022-10-15 - v0.6.5
 
 - Added to `knob_extractor`:
-  - proper state extraction. 
+  - proper state extraction.
   - IP2 and IP8 separation/crossing variables.
 
 #### 2022-10-12 - v0.6.4
@@ -264,7 +279,7 @@
 #### 2022-05-19 - v0.3.0 - _jdilly_
 
 - Added:
-  - Linfile cleaning script. 
+  - Linfile cleaning script.
 
 #### 2022-04-25 - v0.2.7 - _awegshe_
 
@@ -359,7 +374,7 @@
   - Spectrum Plotting
   - Turn-by-Turn Converter
 
-- `setup.py` and packaging functionality 
+- `setup.py` and packaging functionality
 - Automated CI
   - Multiple versions of python
   - Accuracy tests

diff --git a/doc/index.rst b/doc/index.rst
@@ -24,6 +24,21 @@ Package Reference
 
    modules/*
 
+Citing
+======
+
+If ``omc3`` has been significant in your work, and you would like to acknowledge the package in your academic publication, please consider citing the following:
+
+.. code-block:: bibtex
+
+   @software{omc3,
+   author    = {OMC-Team and Malina, L and Dilly, J and Hofer, M and Soubelet, F and Wegscheider, A and Coello De Portugal, J and Le Garrec, M and Persson, T and Keintzel, J and Garcia Morales, H and Tomás, R},
+   doi       = {10.5281/zenodo.5705625},
+   publisher = {Zenodo},
+   title     = {OMC3},
+   url       = {https://doi.org/10.5281/zenodo.5705625},
+   year      = 2022
+   }
 
 Indices and tables
 ==================

diff --git a/omc3/__init__.py b/omc3/__init__.py
@@ -11,7 +11,7 @@
 __title__ = "omc3"
 __description__ = "An accelerator physics tools package for the OMC team at CERN."
 __url__ = "https://github.com/pylhc/omc3"
-__version__ = "0.16.0"
+__version__ = "0.16.1"
 __author__ = "pylhc"
 __author_email__ = "[email protected]"
 __license__ = "MIT"

diff --git a/omc3/correction/response_io.py b/omc3/correction/response_io.py
@@ -4,104 +4,134 @@
 
 Input and output functions for response matrices.
 """
+
+from __future__ import annotations
+
+import logging
+import warnings
 from collections import defaultdict
+from collections.abc import Sequence
+from contextlib import contextmanager
 from pathlib import Path
-from typing import Dict, Sequence, Set
 
 import pandas as pd
-import logging
-
+from tables import NaturalNameWarning
 
 LOG = logging.getLogger(__name__)
 
-COMPLIB = 'blosc'  # zlib is the standard compression
-COMPLEVEL = 9  # goes from 0-9, 9 is highest compression, None deactivates if COMPLIB is None
+COMPLIB: str = "blosc"  # zlib is the standard compression
+COMPLEVEL: int = 9  # goes from 0-9, 9 is highest compression, None deactivates if COMPLIB is None
 
 
 # Fullresponse -----------------------------------------------------------------
 
 
-def read_fullresponse(path: Path, optics_parameters: Sequence[str] = None) -> Dict[str, pd.DataFrame]:
+def read_fullresponse(path: Path, optics_parameters: Sequence[str] = None) -> dict[str, pd.DataFrame]:
     """Load the response matrices from disk.
     Beware: As empty DataFrames are skipped on write,
     default for not found entries are empty DataFrames.
     """
     if not path.exists():
         raise IOError(f"Fullresponse file {str(path)} does not exist.")
-
     LOG.info(f"Loading response matrices from file '{str(path)}'")
-    with pd.HDFStore(path, mode='r') as store:
-        _check_keys(store, optics_parameters, 'fullresponse')
-
-        fullresponse = defaultdict(pd.DataFrame)
-        if optics_parameters is None:
-            optics_parameters = _main_store_groups(store)
-        for param in optics_parameters:
-            fullresponse[param] = store[param]
+
+    # If encountering issues, remove the context manager and debug
+    with ignore_natural_name_warning():
+        with pd.HDFStore(path, mode="r") as store:
+            _check_keys(store, optics_parameters, "fullresponse")
+
+            fullresponse = defaultdict(pd.DataFrame)
+            if optics_parameters is None:
+                optics_parameters = _main_store_groups(store)
+            for param in optics_parameters:
+                fullresponse[param] = store[param]
     return fullresponse
 
 
-def write_fullresponse(path: Path, fullresponse: Dict[str, pd.DataFrame]):
+def write_fullresponse(path: Path, fullresponse: dict[str, pd.DataFrame]):
     """Write the full response matrices to disk.
     Beware: Empty Dataframes are skipped! (HDF creates gigantic files otherwise)
     """
     LOG.info(f"Saving response matrices into file '{str(path)}'")
     if path.exists():
         LOG.warning(f"Fullresponse file {str(path)} already exist and will be overwritten.")
-    with pd.HDFStore(path, mode='w', complib=COMPLIB, complevel=COMPLEVEL) as store:
-        for param, response_df in fullresponse.items():
-            store.put(value=response_df, key=param, format="table")
+
+    # If encountering issues, remove the context manager and debug
+    with ignore_natural_name_warning():
+        with pd.HDFStore(path, mode="w", complib=COMPLIB, complevel=COMPLEVEL) as store:
+            for param, response_df in fullresponse.items():
+                store.put(value=response_df, key=param, format="table")
 
 
 # Varmap -----------------------------------------------------------------------
 
 
-def read_varmap(path: Path, k_values: Sequence[str] = None) -> Dict[str, Dict[str, pd.Series]]:
+def read_varmap(path: Path, k_values: Sequence[str] = None) -> dict[str, dict[str, pd.Series]]:
     """Load the variable mapping file from disk.
     Beware: As empty DataFrames are skipped on write,
     default for not found entries are empty Series.
     """
     if not path.exists():
         raise IOError(f"Varmap file {str(path)} does not exist.")
-
     LOG.info(f"Loading varmap from file '{str(path)}'")
-    with pd.HDFStore(path, mode='r') as store:
-        _check_keys(store, k_values, 'varmap')
-
-        varmap = defaultdict(lambda: defaultdict(pd.Series))
-        for key in store.keys():
-            _, param, subparam = key.split('/')
-            if k_values is not None and param not in k_values:
-                continue
-            varmap[param][subparam] = store[key]
+
+    # If encountering issues, remove the context manager and debug
+    with ignore_natural_name_warning():
+        with pd.HDFStore(path, mode="r") as store:
+            _check_keys(store, k_values, "varmap")
+
+            varmap = defaultdict(lambda: defaultdict(pd.Series))
+            for key in store.keys():
+                _, param, subparam = key.split("/")
+                if k_values is not None and param not in k_values:
+                    continue
+                varmap[param][subparam] = store[key]
     return varmap
 
 
-def write_varmap(path: Path, varmap: Dict[str, Dict[str, pd.Series]]):
+def write_varmap(path: Path, varmap: dict[str, dict[str, pd.Series]]):
     """Write the  variable mapping file to disk.
     Beware: Empty Dataframes are skipped! (HDF creates gigantic files otherwise)
     """
     LOG.info(f"Saving varmap into file '{str(path)}'")
-    with pd.HDFStore(path, mode='w', complib=COMPLIB, complevel=COMPLEVEL) as store:
-        for param, sub in varmap.items():
-            for subparam, varmap_series in sub.items():
-                store.put(value=varmap_series, key=f"{param}/{subparam}", format="table")
+
+    # If encountering issues, remove the context manager and debug
+    with ignore_natural_name_warning():
+        with pd.HDFStore(path, mode="w", complib=COMPLIB, complevel=COMPLEVEL) as store:
+            for param, sub in varmap.items():
+                for subparam, varmap_series in sub.items():
+                    store.put(value=varmap_series, key=f"{param}/{subparam}", format="table")
 
 
 # Helper -----------------------------------------------------------------------
 
 
-def _check_keys(store: pd.HDFStore, keys: Sequence[str], id:str):
+def _check_keys(store: pd.HDFStore, keys: Sequence[str], id: str):
     if keys is None:
         return
 
     groups = _main_store_groups(store)
     not_found = [k for k in keys if k not in groups]
     if len(not_found):
-        raise ValueError(f"The following keys could not be found in {id} file:"
-                         f" {', '.join(not_found)}")
+        raise ValueError(f"The following keys could not be found in {id} file:" f" {', '.join(not_found)}")
 
 
-def _main_store_groups(store: pd.HDFStore) -> Set[str]:
+def _main_store_groups(store: pd.HDFStore) -> set[str]:
     """Returns sequence of unique main store groups."""
-    return {k.split('/')[1] for k in store.keys()}
+    return {k.split("/")[1] for k in store.keys()}
+
+
+@contextmanager
+def ignore_natural_name_warning():
+    """
+    This context manager catches and ignores the 'NaturalNameWarning'
+    emitted within, which is our case comes from pytables. It warns about
+    table entries such as 'kq4.r8b2' which we can't access with syntax such
+    as some_table.kq4.r8b2 but we don't care about this. We let pandas handle
+    the access with getattr (which works).
+
+    If encountering issues, comment out the context manager and debug.
+    """
+    with warnings.catch_warnings():
+        warnings.simplefilter("ignore", NaturalNameWarning)
+        yield
diff --git a/omc3/definitions/formats.py b/omc3/definitions/formats.py
@@ -5,7 +5,7 @@
 Recurring formats are defined here.
 """
 from pathlib import Path
-from datetime import datetime
+from datetime import datetime, timezone
 
 TIME = "%Y_%m_%d@%H_%M_%S_%f"  # CERN default
 CONFIG_FILENAME = "{script:s}_{time:s}.ini"
@@ -16,6 +16,6 @@ def get_config_filename(script):
     """Default Filename for config-files. Call from script with ``__file__``."""
     return CONFIG_FILENAME.format(
         script=Path(script).name.split('.')[0],
-        time=datetime.utcnow().strftime(TIME)
+        time=datetime.now(timezone.utc).strftime(TIME)
     )