Chore: remove unaltered python

.gitignore

@@ -459,3 +459,9 @@ dask-worker-space/
 
 # Others
 src/main.*.cpp
+
+# Unaltered LightGBM python modules fetched at build time (see python-package/setup.py)
+python-package/fairgbm/callback.py
+python-package/fairgbm/libpath.py
+python-package/fairgbm/plotting.py
+python-package/fairgbm/engine.py

python-package/fairgbm/basic.py

@@ -155,11 +155,11 @@ def list_to_1d_numpy(data, dtype=np.float32, name='list'):
         array = data.ravel()
         return cast_numpy_1d_array_to_dtype(array, dtype)
     elif is_1d_list(data):
-        return np.array(data, dtype=dtype, copy=False)
+        return np.asarray(data, dtype=dtype)
     elif isinstance(data, pd_Series):
         if _get_bad_pandas_dtypes([data.dtypes]):
             raise ValueError('Series.dtypes must be int, float or bool')
-        return np.array(data, dtype=dtype, copy=False)  # SparseArray should be supported as well
+        return np.asarray(data, dtype=dtype)  # SparseArray should be supported as well
     else:
         raise TypeError("Wrong type({0}) for {1}.\n"
                         "It should be list, numpy 1-D array or pandas Series".format(type(data).__name__, name))
@@ -456,7 +456,7 @@ def convert_from_sliced_object(data):
 def c_float_array(data):
     """Get pointer of float numpy array / list."""
     if is_1d_list(data):
-        data = np.array(data, copy=False)
+        data = np.asarray(data)
     if is_numpy_1d_array(data):
         data = convert_from_sliced_object(data)
         assert data.flags.c_contiguous
@@ -477,7 +477,7 @@ def c_float_array(data):
 def c_int_array(data):
     """Get pointer of int numpy array / list."""
     if is_1d_list(data):
-        data = np.array(data, copy=False)
+        data = np.asarray(data)
     if is_numpy_1d_array(data):
         data = convert_from_sliced_object(data)
         assert data.flags.c_contiguous
@@ -721,7 +721,7 @@ def predict(self, data, start_iteration=0, num_iteration=-1,
                 lines = f.readlines()
                 nrow = len(lines)
                 preds = [float(token) for line in lines for token in line.split('\t')]
-                preds = np.array(preds, dtype=np.float64, copy=False)
+                preds = np.asarray(preds, dtype=np.float64)
         elif isinstance(data, scipy.sparse.csr_matrix):
             preds, nrow = self.__pred_for_csr(data, start_iteration, num_iteration, predict_type)
         elif isinstance(data, scipy.sparse.csc_matrix):
@@ -778,7 +778,7 @@ def __pred_for_np2d(self, mat, start_iteration, num_iteration, predict_type):
 
         def inner_predict(mat, start_iteration, num_iteration, predict_type, preds=None):
             if mat.dtype == np.float32 or mat.dtype == np.float64:
-                data = np.array(mat.reshape(mat.size), dtype=mat.dtype, copy=False)
+                data = np.asarray(mat.reshape(mat.size), dtype=mat.dtype)
             else:  # change non-float data to float data, need to copy
                 data = np.array(mat.reshape(mat.size), dtype=np.float32)
             ptr_data, type_ptr_data, _ = c_float_array(data)
@@ -1312,7 +1312,7 @@ def __init_from_np2d(self, mat, params_str, ref_dataset):
 
         self.handle = ctypes.c_void_p()
         if mat.dtype == np.float32 or mat.dtype == np.float64:
-            data = np.array(mat.reshape(mat.size), dtype=mat.dtype, copy=False)
+            data = np.asarray(mat.reshape(mat.size), dtype=mat.dtype)
         else:  # change non-float data to float data, need to copy
             data = np.array(mat.reshape(mat.size), dtype=np.float32)
 
@@ -1350,7 +1350,7 @@ def __init_from_list_np2d(self, mats, params_str, ref_dataset):
             nrow[i] = mat.shape[0]
 
             if mat.dtype == np.float32 or mat.dtype == np.float64:
-                mats[i] = np.array(mat.reshape(mat.size), dtype=mat.dtype, copy=False)
+                mats[i] = np.asarray(mat.reshape(mat.size), dtype=mat.dtype)
             else:  # change non-float data to float data, need to copy
                 mats[i] = np.array(mat.reshape(mat.size), dtype=np.float32)
 

python-package/fairgbm/compat.py

@@ -86,8 +86,30 @@ def _check_sample_weight(sample_weight, X, dtype=None):
     LGBMNotFittedError = NotFittedError
     _LGBMStratifiedKFold = StratifiedKFold
     _LGBMGroupKFold = GroupKFold
-    _LGBMCheckXY = check_X_y
-    _LGBMCheckArray = check_array
+
+    # scikit-learn >= 1.6 deprecated the ``force_all_finite`` parameter of the
+    # validation helpers in favour of ``ensure_all_finite`` (removed entirely in
+    # 1.8). Wrap the validators so FairGBM's ``force_all_finite=...`` call sites
+    # keep working across scikit-learn versions.
+    import inspect as _inspect
+
+    def _finite_kwarg_compat(_func):
+        _params = _inspect.signature(_func).parameters
+        if 'force_all_finite' in _params:
+            return _func  # old scikit-learn: pass through unchanged
+
+        def _wrapped(*args, **kwargs):
+            if 'force_all_finite' in kwargs:
+                value = kwargs.pop('force_all_finite')
+                if 'ensure_all_finite' in _params:
+                    kwargs['ensure_all_finite'] = value
+                # else: neither kwarg supported -> drop it silently
+            return _func(*args, **kwargs)
+
+        return _wrapped
+
+    _LGBMCheckXY = _finite_kwarg_compat(check_X_y)
+    _LGBMCheckArray = _finite_kwarg_compat(check_array)
     _LGBMCheckSampleWeight = _check_sample_weight
     _LGBMAssertAllFinite = assert_all_finite
     _LGBMCheckClassificationTargets = check_classification_targets