Add version numbers to public functions

src/pnanolocz/level.py

@@ -7,7 +7,7 @@
 row- or column-wise artefacts commonly observed in AFM topographic data.
 
 All public leveling functions accept an *exclusion mask* (same convention as
-``pnanolocz_lib.thresholder``): ``True`` = excluded, ``False`` = valid.
+``pnanolocz.thresholder``): ``True`` = excluded, ``False`` = valid.
 Excluded pixels are omitted from fitting using MATLAB-style NaN-outside
 semantics (i.e., excluded pixels behave like NaN during fitting) but are
 preserved in the output array.
@@ -63,10 +63,10 @@
 
 Examples
 --------
->>> from pnanolocz_lib.level import level
+>>> from pnanolocz.level import level
 >>> leveled_stack = apply_level(stack, polyx=2, polyy=2, method="plane")
 
->>> from pnanolocz_lib.level import level_plane
+>>> from pnanolocz.level import level_plane
 >>> flattened = level_plane(img, mask=None, polyx=2, polyy=2)
 
 Authors
@@ -845,6 +845,10 @@ def apply_level(
     operations (e.g. ``where(valid, value, nan)``) during fitting.
     - Method-specific MATLAB parity notes (e.g., stage gating such as `polyy > 0`
     in ``"line"``) are documented in the corresponding function docstrings.
+
+    Version
+    -------
+    0.1.0
     """
     img = np.asarray(img)
     is_stack = img.ndim == 3
@@ -921,6 +925,9 @@ def apply_level(
     return np.asarray(result) if is_stack else np.asarray(result[0])
 
 
+apply_level.__version__ = "0.1.0"  # type: ignore[attr-defined]
+
+
 def get_background(
     img: np.ndarray[Any, np.dtype[np.float64]],
     polyx: int,
@@ -982,6 +989,10 @@ def get_background(
     certain parameter values) are inherited here.
     - Excluded pixels are preserved in the output arrays; masking primarily affects
     which pixels contribute to fitted estimates.
+
+    Version
+    -------
+    0.1.0
     """
     img = np.asarray(img)
     is_stack = img.ndim == 3
@@ -1028,6 +1039,9 @@ def get_background(
     return np.asarray(result) if is_stack else np.asarray(result[0])
 
 
+get_background.__version__ = "0.1.0"  # type: ignore[attr-defined]
+
+
 __all__ = [
     "apply_level",
     "level_plane",
@@ -1037,4 +1051,5 @@ def get_background(
     "level_smed_line",
     "level_mean_plane",
     "level_log_y",
+    "get_background",
 ]

src/pnanolocz/level_auto.py

@@ -972,6 +972,10 @@ def apply_level_auto(
       ``_compute_gauss_limits`` when a thresholder step declares ``args=['gauss_*']``.
     - The Gaussian-derived histogram bounds are calculated for each frame rather
       than globally across the stack, diverging from MATLAB behavior.
+
+    Version
+    -------
+    0.1.0
     """
     img_stack = np.asarray(img_stack)
     if img_stack.ndim == 2:
@@ -1041,4 +1045,7 @@ def apply_level_auto(
     return np.asarray(result[0]) if was_2d else np.asarray(result)
 
 
+apply_level_auto.__version__ = "0.1.0"  # type: ignore[attr-defined]
+
+
 __all__ = ["apply_level_auto", "ROUTINES"]

src/pnanolocz/level_weighted.py

@@ -8,7 +8,7 @@
 and non-uniform masking effects.
 
 All public leveling functions accept an *exclusion mask* (same convention as
-``pnanolocz_lib.thresholder``): ``True`` = excluded, ``False`` = valid. Excluded
+``pnanolocz.thresholder``): ``True`` = excluded, ``False`` = valid. Excluded
 pixels are omitted from region formation and fitting using MATLAB-style NaN-outside
 semantics (i.e., excluded pixels behave like NaN during fitting) but are
 preserved in the output array.
@@ -55,7 +55,7 @@
 
 Examples
 --------
->>> from pnanolocz_lib.level_weighted import apply_level_weighted
+>>> from pnanolocz.level_weighted import apply_level_weighted
 >>> leveled = apply_level_weighted(img, polyx=2, polyy=1, method='plane', mask=mask)
 
 Authors
@@ -849,6 +849,10 @@ def apply_level_weighted(
       ``bwconncomp(mask, 8)``.
     - A MATLAB-style minimum region area is enforced via:
       ``min_area = max(1, floor(0.01 * H * W))``.
+
+    Version
+    -------
+    0.1.0
     """
     arr = np.asarray(img, dtype=np.float64)
     is_stack = arr.ndim == 3
@@ -897,6 +901,9 @@ def apply_level_weighted(
     return np.asarray(stacked if is_stack else stacked[0])
 
 
+apply_level_weighted.__version__ = "0.1.0"  # type: ignore[attr-defined]
+
+
 __all__ = [
     "apply_level_weighted",
     "level_weighted_plane",

src/pnanolocz/thresholder.py

@@ -1027,6 +1027,10 @@ def apply_thresholder(
     >>> # Otsu on a stack, then invert to obtain a validity mask (True = valid)
     >>> excl = apply_thresholder(stack, method="otsu")
     >>> valid = np.logical_not(excl)
+
+    Version
+    -------
+    0.1.0
     """
     method = method.lower()
     if method not in _METHOD_MAP:
@@ -1069,6 +1073,9 @@ def apply_thresholder(
     return result
 
 
+apply_thresholder.__version__ = "0.1.0"  # type: ignore[attr-defined]
+
+
 __all__ = [
     "apply_thresholder",
     "selection",