Refactor molecule fitting

OES_toolbox/Widgets.py

@@ -6,8 +6,13 @@
 import pyqtgraph as pg
 import qtawesome as qta
 
+import Moose
+
 from OES_toolbox.file_handling import FileLoader, SpectraDataset
 from OES_toolbox.logger import Logger
+from OES_toolbox.lazy_import import lazy_import
+
+pd = lazy_import("pandas")
 
 from typing import TYPE_CHECKING
 from collections.abc import Callable
@@ -410,3 +415,67 @@ def _populate_with_data(self, dataset:SpectraDataset, label=None):
             self.is_content = True # needed to force non-nested files to plot their data, without adding a child node.
             self.set_spectrum(x,y,shift=shift,bg=dataset.background,name=None,bg_is_internal=dataset.has_background)
         self.set_background(None)
+
+class MoleculeCheckBox(QCheckBox):
+    """A checkbox that enables a user to select molecular bands to show or fit.
+
+    Handles both static spectra from LIFBASE simulation, as well as database files for simulation/fitting.
+
+    Only loads data when needed, caching it to the `_db` attribute.
+
+    A subset of the data (in wavelength range), can be obtained using the `get_db` method.
+    """
+
+    def __init__(self, ident: str, label: str = None, src: str = "LIFBASE", parent=None):
+        super().__init__(parent=parent)
+        self.ident = ident
+        self.label = ident if label is None else label
+        self.src = src
+        self.can_fit = self.src == "mOES"
+        self.setText(self.label)
+        self._db = None
+
+    @property
+    def db(self):
+        if self._db is None:
+            self._load_database()
+        return self._db
+
+    def _load_database(self):
+        """Load a database or sample LIFBASE simulation and cache it on the instance.
+
+        The database/file is only loaded if the `_db` attribute is (still) `None`.
+
+        No filtering/slicing is applied to the database, so it can be cached and re-used effectively.
+        """
+        if self.can_fit and self._db is None:
+            with pg.ProgressDialog(f"Loading line-by-line database: {self.label}",cancelText=None,wait=0,busyCursor=True) as _diag:
+                self._db = Moose.query_DB(self.ident)
+        elif not self.can_fit and self.src=="LIFBASE" and self._db is None:
+            with pg.ProgressDialog(f"Loading LIFBASE simulation: {self.label}",cancelText=None, wait=0) as _diag:
+                # Read the LIFBASE output files, which use lots of whitespace padding and must thus be coerced to float rather than string
+                # FileLoader._read_generic_text fails correctly detecting separator/delimiter because of whitespacing.
+                file_path = f"{Path(__file__).parent}/data/mol_spec/{self.ident}.mod"
+                data = pd.read_csv(file_path,header=None, sep=",",decimal='.',dtype=float,names=["wl","I"])
+                data.wl=data.iloc[:,0]/10 # Angstrom to nm
+                self._db = data
+
+    def get_db(self, wavelength_interval:tuple[float,float]|None = None, wl_pad:float|int = 10):
+        """Return a slice of a database, or LIFBASE spectrum, within the specified `wavelength_interval`.
+
+        If no interval is provided (the default), it will lookup the active bounds from the main window.
+
+        If no data(base) has been loaded yet, calling this method will trigger a `load_database` call, which will cache the data.
+
+        To avoid edge effects, it will look up a slightly wider slice of data on both sides of the interval, based on the `wl_pad` argument (default=10).
+
+        Note: for automatic lookup of the bounds to work correctly, the widget instance must have a parent that is part of the OESToolbox main window.
+        """
+        colname = 'wl' if self.src.upper()=="LIFBASE" else 'air_wavelength'
+        if wavelength_interval is None:
+            wl_min, wl_max, *_= self.window().get_bounds()
+        else:
+            wl_min,wl_max = wavelength_interval
+        data = self.db[self.db[colname].between(wl_min - wl_pad, wl_max + wl_pad)]
+        return data
+

OES_toolbox/ident.py

@@ -71,18 +71,8 @@ def update_spec_ident(self):
         self.mw.ident_table.setRowCount(0)
 
         # find wavelength range from file spec
-        for plot_item in self.mw.specplot.listDataItems():
-            if "file" in plot_item.name():
-                x0, x1 = plot_item.dataBounds(0)
-                if lim_unset:
-                    min_x = x0
-                    max_x = x1
-                    max_y = plot_item.dataBounds(1)[1]
-                    lim_unset = False
-
-                min_x = min(min_x, x0)
-                max_x = max(max_x, x1)
-
+        min_x, max_x, min_y, max_y = self.mw.get_bounds()
+
         # fetch options
         if self.mw.ident_int_cbox.currentIndex() == 1:
             Te = self.mw.ident_Te.value()