Use ones_like

examples/linear-model.py

@@ -17,7 +17,7 @@
 import ase.io
 import numpy as np
 from equistore import Labels
-from equistore.operations import slice, sum_over_samples
+from equistore.operations import slice, sum_over_samples, ones_like
 from rascaline import SoapPowerSpectrum
 
 from equisolve.numpy.models.linear_model import Ridge
@@ -132,9 +132,8 @@
 #
 # For this we create a TensorMap containing with the desired regulerizer
 
-
-alpha_dict = {"values": 1e-5}
-alpha = dictionary_to_tensormap(alpha_dict, X)
+alpha = ones_like(X)
+alpha.block().values[:] *= 1e-5
 
 # %%
 #
@@ -173,8 +172,8 @@
 # Next we create a sample weighting :class:`equistiore.TensorMap` that weights energies
 # five times more then the forces.
 
-sw_dict = {"values": 5, "positions": 1}
-sw = dictionary_to_tensormap(sw_dict, y)
+sw = ones_like(y)
+sw.block().values[:] *= 5
 
 # %%
 #

src/equisolve/utils/__init__.py

@@ -6,13 +6,12 @@
 # Released under the BSD 3-Clause "New" or "Revised" License
 # SPDX-License-Identifier: BSD-3-Clause
 
-from .convert import ase_to_tensormap, dictionary_to_tensormap, properties_to_tensormap
+from .convert import ase_to_tensormap, properties_to_tensormap
 from .metrics import rmse
 
 
 __all__ = [
     "ase_to_tensormap",
-    "dictionary_to_tensormap",
     "properties_to_tensormap",
     "rmse",
 ]

src/equisolve/utils/convert.py

@@ -141,34 +141,3 @@ def properties_to_tensormap(
         )
 
     return TensorMap(Labels.single(), [block])
-
-
-def dictionary_to_tensormap(d: dict, ts: TensorMap) -> TensorMap:
-    """TensorMap with data provided by a dictionary and meta data from a TensorMap.
-
-    Parameters which are no keys in ``d`` but present in ``block`` will be copied
-    into the new block.
-
-    Should be moved to equistore!
-
-    :paramas d: Keys of ``d`` have to be the 'parameters' of provided TensorMap i.e.
-                ``'values'``, ``'positions'``, ``'cell'``. Values of ``d`` can either
-                be scalars or must have the same shape as the data associated with
-                each 'parameter'.
-    :params block: Reference :class:`equistore.TensorMap` for gathering the
-                   meta data
-    :returns ts_d: block containing the data of ``d`` with the metadata of ``block``.
-    """
-    blocks = []
-    for block in ts.blocks():
-        new_block = block.copy()
-
-        for parameter, data in d.items():
-            if parameter == "values":
-                new_block.values[:] = data
-            else:
-                new_block.gradient(parameter).data[:] = data
-
-        blocks.append(new_block)
-
-    return TensorMap(ts.keys, blocks)