splusdata 5.49__tar.gz → 5.51__tar.gz

{splusdata-5.49/splusdata.egg-info → splusdata-5.51}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: splusdata
-Version: 5.49
+Version: 5.51
 Summary: Download SPLUS catalogs, FITS and more
 Author-email: Gustavo Schwarz <gustavo.b.schwarz@gmail.com>
 License: Apache-2.0
@@ -10,7 +10,7 @@ Classifier: License :: OSI Approved :: Apache Software License
 Requires-Python: >=3.7
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: adss>=1.36
+Requires-Dist: adss>=1.38
 Requires-Dist: requests
 Requires-Dist: astropy
 Requires-Dist: astroquery

{splusdata-5.49 → splusdata-5.51}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "splusdata"
-version = "5.49"
+version = "5.51"
 description = "Download SPLUS catalogs, FITS and more"
 authors = [
     { name = "Gustavo Schwarz", email = "gustavo.b.schwarz@gmail.com" }
@@ -18,7 +18,7 @@ classifiers = [
     "License :: OSI Approved :: Apache Software License"
 ]
 dependencies = [
-    "adss>=1.36",
+    "adss>=1.38",
     "requests",
     "astropy",
     "astroquery",

splusdata-5.51/splusdata/features/zeropoints/zp_map.py

@@ -0,0 +1,271 @@
+from scipy.interpolate import RegularGridInterpolator
+import numpy as np
+
+import numpy as np
+import warnings
+from scipy.interpolate import RegularGridInterpolator
+
+def _reconstruct_centers_from_model(model, axis="ra", grid_len=None):
+    """
+    Rebuild centers to match the (possibly padded) grid length using the model's
+    ra_min/ra_max/dec_min/dec_max, bins, and padding.
+
+    The model saved edges with length `bins`, which yields B = bins-1 actual bins
+    before padding. After padding, the grid dimension is B + 2*padding.
+    Centers are midpoints of each (possibly padded) bin.
+
+    Parameters
+    ----------
+    model : dict
+    axis : str
+        "ra" or "dec"
+    grid_len : int
+        Target number of centers (should equal grid.shape[dim])
+
+    Returns
+    -------
+    np.ndarray
+    """
+    assert axis in ("ra", "dec")
+    amin = model[f"{axis}_min"]
+    amax = model[f"{axis}_max"]
+    bins = int(model.get("bins", 15))
+    padding = int(model.get("padding", 1))
+
+    # B is the number of *actual* bins before padding (because edges length = bins)
+    B = bins - 1
+    if B <= 0:
+        raise ValueError(f"Invalid bins in model: bins={bins}")
+
+    # native bin width (pre-padding)
+    dA = (amax - amin) / B
+
+    # original centers (length B)
+    centers_core = amin + (np.arange(B) + 0.5) * dA
+
+    # padded centers: extend by `padding` bins on both sides at same spacing
+    if padding > 0:
+        left = centers_core[0] - dA * np.arange(padding, 0, -1)
+        right = centers_core[-1] + dA * np.arange(1, padding + 1)
+        centers = np.concatenate([left, centers_core, right])
+    else:
+        centers = centers_core
+
+    if grid_len is not None and len(centers) != grid_len:
+        # If still mismatched, resample linearly across the padded span as a last resort.
+        warnings.warn(
+            f"{axis.upper()} centers length ({len(centers)}) != grid axis length ({grid_len}). "
+            "Resampling centers to match grid shape."
+        )
+        # Rebuild centers uniformly across the total span of the padded grid:
+        total_span = (amax - amin) + 2 * padding * dA
+        a_min_padded = amin - padding * dA
+        centers = a_min_padded + (np.arange(grid_len) + 0.5) * (total_span / grid_len)
+
+    return centers
+
+def zp_at_coord(model, ra, dec, margin=0.1):
+    """
+    Get zero-point correction(s) for coordinate(s).
+
+    Accepts:
+      - ra:  float or array-like
+      - dec: float or array-like
+
+    Supports broadcasting:
+      - scalar ra with array dec
+      - array ra with scalar dec
+      - array ra with array dec (same shape or broadcastable)
+
+    Returns:
+      - float if both inputs are scalar
+      - np.ndarray otherwise
+
+    Notes:
+      - Keeps your original behavior: if ANY point is out of bounds (with margin)
+        or interpolation yields NaN, it raises Exception (instead of partial fill).
+    """
+    global_median = float(model.get("global_median", 0.0))
+
+    # If no grid info, fallback
+    if not ("grid" in model and "ra_centers" in model and "dec_centers" in model):
+        if np.isscalar(ra) and np.isscalar(dec):
+            return global_median
+        ra_arr = np.asarray(ra, dtype=float)
+        dec_arr = np.asarray(dec, dtype=float)
+        ra_b, dec_b = np.broadcast_arrays(ra_arr, dec_arr)
+        return np.full(ra_b.shape, global_median, dtype=float)
+
+    # Load saved arrays
+    grid = np.asarray(model["grid"], dtype=float)
+    ra_centers = np.asarray(model.get("ra_centers", []), dtype=float)
+    dec_centers = np.asarray(model.get("dec_centers", []), dtype=float)
+
+    # Rebuild centers if needed (mismatch / empty)
+    need_rebuild = (
+        ra_centers.size == 0 or
+        dec_centers.size == 0 or
+        grid.ndim != 2 or
+        grid.shape[0] != ra_centers.size or
+        grid.shape[1] != dec_centers.size
+    )
+    if need_rebuild:
+        ra_centers = _reconstruct_centers_from_model(model, "ra", grid_len=grid.shape[0])
+        dec_centers = _reconstruct_centers_from_model(model, "dec", grid_len=grid.shape[1])
+
+    # Normalize inputs + broadcast
+    ra_is_scalar = np.isscalar(ra)
+    dec_is_scalar = np.isscalar(dec)
+
+    ra_arr = np.asarray(ra, dtype=float)
+    dec_arr = np.asarray(dec, dtype=float)
+    ra_b, dec_b = np.broadcast_arrays(ra_arr, dec_arr)
+
+    # Vectorized bounds check
+    ra_min, ra_max = float(np.min(ra_centers)), float(np.max(ra_centers))
+    dec_min, dec_max = float(np.min(dec_centers)), float(np.max(dec_centers))
+
+    in_bounds = (
+        (ra_b >= (ra_min - margin)) & (ra_b <= (ra_max + margin)) &
+        (dec_b >= (dec_min - margin)) & (dec_b <= (dec_max + margin))
+    )
+
+    if not np.all(in_bounds):
+        bad = np.argwhere(~in_bounds)
+        i0 = tuple(bad[0])  # first offending index
+        warnings.warn(
+            f"Some coordinates are outside the grid range RA=[{ra_min:.3f}, {ra_max:.3f}] "
+            f"Dec=[{dec_min:.3f}, {dec_max:.3f}] (margin={margin:.3f}). "
+            f"Example at index {i0}: (RA={ra_b[i0]:.3f}, Dec={dec_b[i0]:.3f}). "
+            "Falling back to global median (raising exception, per original behavior)."
+        )
+        raise Exception("Some coordinates are outside the grid range.")
+
+    # Build interpolator once
+    interpolator = RegularGridInterpolator(
+        (ra_centers, dec_centers),
+        grid,
+        bounds_error=False,
+        fill_value=np.nan,
+    )
+
+    # Interpolate all points
+    pts = np.column_stack([ra_b.ravel(), dec_b.ravel()])  # (N, 2)
+    zp = interpolator(pts).reshape(ra_b.shape)
+
+    if np.any(np.isnan(zp)):
+        bad = np.argwhere(np.isnan(zp))
+        i0 = tuple(bad[0])
+        warnings.warn(
+            f"Interpolation failed (NaN) for some coordinates. "
+            f"Example at index {i0}: (RA={ra_b[i0]:.3f}, Dec={dec_b[i0]:.3f}). "
+            "Returning global median (raising exception, per original behavior)."
+        )
+        raise Exception("Interpolation failed (NaN) for some coordinates.")
+
+    out = zp + global_median
+
+    # Return scalar if scalar inputs
+    if ra_is_scalar and dec_is_scalar:
+        return float(out)
+
+    return out
+    """
+    Vectorized zero-point correction lookup.
+
+    Accepts:
+      - ra: float or array-like
+      - dec: float or array-like
+    Supports broadcasting:
+      - ra scalar + dec array
+      - ra array + dec scalar
+      - ra array + dec array (same shape or broadcastable)
+
+    Returns:
+      - float if both inputs are scalar (and return_scalar_if_scalar=True)
+      - np.ndarray otherwise
+    """
+    global_median = float(model.get("global_median", 0.0))
+
+    # If no grid, always fallback
+    if not ("grid" in model and ("ra_centers" in model or True) and ("dec_centers" in model or True)):
+        # keep original behavior: return global median only
+        if np.isscalar(ra) and np.isscalar(dec) and return_scalar_if_scalar:
+            return global_median
+        ra_arr = np.asarray(ra, dtype=float)
+        dec_arr = np.asarray(dec, dtype=float)
+        ra_b, dec_b = np.broadcast_arrays(ra_arr, dec_arr)
+        return np.full(ra_b.shape, global_median, dtype=float)
+
+    # Load arrays
+    grid = np.asarray(model["grid"], dtype=float)
+    ra_centers = np.asarray(model.get("ra_centers", []), dtype=float)
+    dec_centers = np.asarray(model.get("dec_centers", []), dtype=float)
+
+    # Rebuild centers if needed
+    need_rebuild = (
+        ra_centers.size == 0 or
+        dec_centers.size == 0 or
+        grid.ndim != 2 or
+        grid.shape[0] != ra_centers.size or
+        grid.shape[1] != dec_centers.size
+    )
+    if need_rebuild:
+        ra_centers = _reconstruct_centers_from_model(model, "ra", grid_len=grid.shape[0])
+        dec_centers = _reconstruct_centers_from_model(model, "dec", grid_len=grid.shape[1])
+
+    # Interpolator (build once)
+    interpolator = RegularGridInterpolator(
+        (ra_centers, dec_centers),
+        grid,
+        bounds_error=False,
+        fill_value=np.nan,
+    )
+
+    # Normalize inputs to arrays + broadcast
+    ra_is_scalar = np.isscalar(ra)
+    dec_is_scalar = np.isscalar(dec)
+
+    ra_arr = np.asarray(ra, dtype=float)
+    dec_arr = np.asarray(dec, dtype=float)
+    ra_b, dec_b = np.broadcast_arrays(ra_arr, dec_arr)
+
+    # Bounds check (vectorized) with margin
+    ra_min, ra_max = float(np.min(ra_centers)), float(np.max(ra_centers))
+    dec_min, dec_max = float(np.min(dec_centers)), float(np.max(dec_centers))
+
+    in_bounds = (
+        (ra_b >= (ra_min - margin)) & (ra_b <= (ra_max + margin)) &
+        (dec_b >= (dec_min - margin)) & (dec_b <= (dec_max + margin))
+    )
+
+    if not np.all(in_bounds):
+        bad = np.argwhere(~in_bounds)
+        i0 = tuple(bad[0])  # first offending index
+        warnings.warn(
+            f"Some coordinates are outside the grid range (margin={margin} deg). "
+            f"Example at index {i0}: (RA={ra_b[i0]:.3f}, Dec={dec_b[i0]:.3f}) "
+            f"outside RA=[{ra_min:.3f}, {ra_max:.3f}], Dec=[{dec_min:.3f}, {dec_max:.3f}]."
+        )
+        raise Exception("Some coordinates are outside the grid range.")
+
+    # Evaluate interpolation for all points
+    pts = np.column_stack([ra_b.ravel(), dec_b.ravel()])  # shape (N, 2)
+    zp = interpolator(pts).reshape(ra_b.shape)
+
+    if np.any(np.isnan(zp)):
+        bad = np.argwhere(np.isnan(zp))
+        i0 = tuple(bad[0])
+        warnings.warn(
+            f"Interpolation returned NaN for some points. "
+            f"Example at index {i0}: (RA={ra_b[i0]:.3f}, Dec={dec_b[i0]:.3f})."
+        )
+        raise Exception("Interpolation failed (NaN) for some points.")
+
+    zp = zp + global_median
+
+    # Return float if scalar inputs
+    if return_scalar_if_scalar and ra_is_scalar and dec_is_scalar:
+        return float(zp)
+
+    return zp

{splusdata-5.49 → splusdata-5.51/splusdata.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: splusdata
-Version: 5.49
+Version: 5.51
 Summary: Download SPLUS catalogs, FITS and more
 Author-email: Gustavo Schwarz <gustavo.b.schwarz@gmail.com>
 License: Apache-2.0
@@ -10,7 +10,7 @@ Classifier: License :: OSI Approved :: Apache Software License
 Requires-Python: >=3.7
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: adss>=1.36
+Requires-Dist: adss>=1.38
 Requires-Dist: requests
 Requires-Dist: astropy
 Requires-Dist: astroquery

{splusdata-5.49 → splusdata-5.51}/splusdata.egg-info/requires.txt

@@ -1,4 +1,4 @@
-adss>=1.36
+adss>=1.38
 requests
 astropy
 astroquery

splusdata-5.49/splusdata/features/zeropoints/zp_map.py

@@ -1,145 +0,0 @@
-from scipy.interpolate import RegularGridInterpolator
-import numpy as np
-
-import numpy as np
-import warnings
-from scipy.interpolate import RegularGridInterpolator
-
-def _reconstruct_centers_from_model(model, axis="ra", grid_len=None):
-    """
-    Rebuild centers to match the (possibly padded) grid length using the model's
-    ra_min/ra_max/dec_min/dec_max, bins, and padding.
-
-    The model saved edges with length `bins`, which yields B = bins-1 actual bins
-    before padding. After padding, the grid dimension is B + 2*padding.
-    Centers are midpoints of each (possibly padded) bin.
-
-    Parameters
-    ----------
-    model : dict
-    axis : str
-        "ra" or "dec"
-    grid_len : int
-        Target number of centers (should equal grid.shape[dim])
-
-    Returns
-    -------
-    np.ndarray
-    """
-    assert axis in ("ra", "dec")
-    amin = model[f"{axis}_min"]
-    amax = model[f"{axis}_max"]
-    bins = int(model.get("bins", 15))
-    padding = int(model.get("padding", 1))
-
-    # B is the number of *actual* bins before padding (because edges length = bins)
-    B = bins - 1
-    if B <= 0:
-        raise ValueError(f"Invalid bins in model: bins={bins}")
-
-    # native bin width (pre-padding)
-    dA = (amax - amin) / B
-
-    # original centers (length B)
-    centers_core = amin + (np.arange(B) + 0.5) * dA
-
-    # padded centers: extend by `padding` bins on both sides at same spacing
-    if padding > 0:
-        left = centers_core[0] - dA * np.arange(padding, 0, -1)
-        right = centers_core[-1] + dA * np.arange(1, padding + 1)
-        centers = np.concatenate([left, centers_core, right])
-    else:
-        centers = centers_core
-
-    if grid_len is not None and len(centers) != grid_len:
-        # If still mismatched, resample linearly across the padded span as a last resort.
-        warnings.warn(
-            f"{axis.upper()} centers length ({len(centers)}) != grid axis length ({grid_len}). "
-            "Resampling centers to match grid shape."
-        )
-        # Rebuild centers uniformly across the total span of the padded grid:
-        total_span = (amax - amin) + 2 * padding * dA
-        a_min_padded = amin - padding * dA
-        centers = a_min_padded + (np.arange(grid_len) + 0.5) * (total_span / grid_len)
-
-    return centers
-
-def zp_at_coord(model, ra, dec, margin=0.1):
-    """
-    Get zero-point correction for a given coordinate.
-
-    Parameters
-    ----------
-    model : dict
-        Zero-point calibration model loaded from JSON.
-    ra : float
-        Right Ascension in degrees.
-    dec : float
-        Declination in degrees.
-    margin : float, optional
-        Allowed margin (in degrees) outside the grid before warning.
-        Default = 0.1 deg.
-
-    Returns
-    -------
-    float
-        Zero-point correction value (mag).
-    """
-    global_median = model.get("global_median", 0.0)
-
-    
-    if "grid" in model and "ra_centers" in model and "dec_centers" in model:
-        ra_centers = np.array(model["ra_centers"])
-        dec_centers = np.array(model["dec_centers"])
-        grid = np.array(model["grid"])
-
-        need_rebuild = (
-            ra_centers.size == 0 or
-            dec_centers.size == 0 or
-            grid.shape[0] != ra_centers.size or
-            grid.shape[1] != dec_centers.size
-        )
-        if need_rebuild:
-            ra_centers = _reconstruct_centers_from_model(model, "ra", grid_len=grid.shape[0])
-            dec_centers = _reconstruct_centers_from_model(model, "dec", grid_len=grid.shape[1])
-            
-        
-        ra_min, ra_max = ra_centers.min(), ra_centers.max()
-        dec_min, dec_max = dec_centers.min(), dec_centers.max()
-
-        # Check bounds with margin
-        if not (ra_min - margin <= ra <= ra_max + margin and
-                dec_min - margin <= dec <= dec_max + margin):
-            warnings.warn(
-                f"Coordinate (RA={ra:.3f}, Dec={dec:.3f}) is outside "
-                f"the grid range RA=[{ra_min:.3f}, {ra_max:.3f}], "
-                f"Dec=[{dec_min:.3f}, {dec_max:.3f}]. "
-                "Falling back to global median."
-            )
-            raise Exception(
-                f"Coordinate (RA={ra}, Dec={dec}) is outside the grid range."
-            )
-
-        # Interpolator
-        interpolator = RegularGridInterpolator(
-            (ra_centers, dec_centers), grid,
-            bounds_error=False,
-            fill_value=np.nan
-        )
-        zp_value = interpolator([[ra, dec]])[0]
-
-        # If interpolator returns NaN, fallback
-        if np.isnan(zp_value):
-            warnings.warn(
-                f"Interpolation failed at (RA={ra:.3f}, Dec={dec:.3f}). "
-                "Returning global median."
-            )
-            raise Exception(
-                f"Interpolation failed at (RA={ra}, Dec={dec}). "
-                "Falling back to global median."
-            )
-
-        return float(zp_value) + global_median
-
-    # Fallback if no grid in model
-    return global_median