Simplify the use of timeslices

.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 1.2.2
+current_version = 1.3.0
 commit = True
 tag = True
 

CITATION.cff

@@ -9,5 +9,5 @@ authors:
       given-names: Adam
 
 title: MUSE_OS
-version: v1.2.2
+version: v1.3.0
 date-released: 2024-08-13

docs/conf.py

@@ -8,7 +8,7 @@
 project = "MUSE"
 copyright = "2024, Imperial College London"
 author = "Imperial College London"
-release = "1.2.2"
+release = "1.3.0"
 version = ".".join(release.split(".")[:2])
 
 # -- General configuration ---------------------------------------------------

docs/inputs/toml.rst

@@ -236,30 +236,6 @@ levels. For instance, there no ``peak`` periods during weekends. All that matter
 that the relative weights (i.e. the number of hours) are consistent and sum up to a
 year.
 
-The input above defines the finest times slice in the code. In order to define rougher
-timeslices we can introduce items in each levels that represent aggregates at that
-level. By default, we have the following:
-
-.. code-block:: TOML
-
-    [timeslices.aggregates]
-    all-day = ["night", "morning", "afternoon", "early-peak", "late-peak", "evening"]
-    all-week = ["weekday", "weekend"]
-    all-year = ["winter", "summer", "spring-autumn"]
-
-Here, ``all-day`` aggregates the full day. However, one could potentially create
-aggregates such as:
-
-.. code-block:: TOML
-
-    [timeslices.aggregates]
-    daylight = ["morning", "afternoon", "early-peak", "late-peak"]
-    nightlife = ["evening", "night"]
-
- It is possible to specify a timeslice level for the mca by adding an
-`mca.timeslice_levels` section, using an inline table format.
-See section on `Timeslices_`.
-
 *outputs_cache*
    This option behaves exactly like `outputs` for sectors and accepts the same options but
    controls the output of cached quantities instead. This option is NOT available for
@@ -423,30 +399,6 @@ Sectors contain a number of subsections:
       Path to a csv file describing the outputs of each technology involved in the sector.
       See :ref:`inputs-iocomms`.
 
-   Once the finest timeslice and its aggregates are given, it is  possible for each sector
-to define the timeslice simply by referring to the slices it will use at each level.
-
-.. _sector-timeslices:
-
-*timeslice_levels*
-   Optional. These define the timeslices of a sector. If not specified, the finest timeslice levels will be used
-   (See `Timeslices`_).
-   It can be implemented with the following rows:
-
-.. code-block:: TOML
-
-    [sectors.some_sector.timeslice_levels]
-    day = ["daylight", "nightlife"]
-    month = ["all-year"]
-
-   Above, ``sectors.some_sector.timeslice_levels.week`` defaults its value in the finest
-   timeslice. Indeed, if the subsection ``sectors.some_sector.timeslice_levels`` is not
-   given, then the sector will default to using the finest timeslices.
-
-   If the MCA uses a rougher
-   timeslice framework, the market will be expressed within it. Hence information from
-   sectors with a finer timeslice framework will be lost.
-
 *subsectors*
 
     Subsectors group together agents into separate groups servicing the demand for
@@ -711,58 +663,3 @@ The following attributes are accepted:
 
       filters.region = ["USA", "ASEA"]
       filters.commodity = ["algae", "fluorescent light"]
-
-
---------------
-Legacy Sectors
---------------
-
-Legacy sectors wrap sectors developed for a previous version of MUSE to the open-source
-version.
-
-Preset sectors are defined in :py:class:`~muse.sectors.PresetSector`.
-
-The can be defined in the TOML file as follows:
-
-.. code-block:: TOML
-
-   [global_input_files]
-   macrodrivers = '{path}/input/Macrodrivers.csv'
-   regions = '{path}/input/Regions.csv'
-   global_commodities = '{path}/input/MUSEGlobalCommodities.csv'
-
-   [sectors.Industry]
-   type = 'legacy'
-   priority = 'demand'
-   agregation_level = 'month'
-   excess = 0
-
-   userdata_path = '{muse_sectors}/Industry'
-   technodata_path = '{muse_sectors}/Industry'
-   timeslices_path = '{muse_sectors}/Industry/TimeslicesIndustry.csv'
-   output_path = '{path}/output'
-
-For historical reasons, the three `global_input_files` above are required. The sector
-itself can use the following attributes.
-
-*type*
-   See the attribute in the standard mode, :ref:`type<sector-type>`. *Legacy* sectors
-   are those with type "legacy".
-
-*priority*
-   See the attribute in the standard mode, :ref:`priority<sector-priority>`.
-
-*agregation_level*
-   Information relevant to the sector's timeslice.
-
-*excess*
-   Excess factor used to model early obsolescence.
-
-*userdata_path*
-   Path to a directory with sector-specific data files.
-
-*technodata_path*
-   Path to a technodata CSV file. See. :ref:`inputs-technodata`.
-
-*output_path*
-   Path to a directory where the sector will write output files.

src/muse/__init__.py

@@ -2,7 +2,7 @@
 
 import os
 
-VERSION = "1.2.2"
+VERSION = "1.3.0"
 
 
 def _create_logger(color: bool = True):
@@ -89,7 +89,6 @@ def add_file_logger() -> None:
     "read_technodictionary",
     "read_technologies",
     "read_timeslice_shares",
-    "read_csv_timeslices",
     "read_settings",
     "read_macro_drivers",
     "read_csv_agent_parameters",

src/muse/__main__.py

@@ -62,5 +62,51 @@ def run():
     muse_main(args.settings, args.model, args.copy)
 
 
+def patched_broadcast_compat_data(self, other):
+    """Patch for xarray.core.variable._broadcast_compat_data.
+
+    This has been introduced to disallow automatic broadcasting along the 'timeslice'
+    dimension.
+
+    If `self` and `other` differ in whether they have a 'timeslice' dimension (in which
+    case automatic broadcasting would normally be performed), an error is raised.
+
+    In this case, developers must explicitly handle broadcasting by calling either
+    `broadcast_timeslice` or `distribute_timeslice` (see `muse.timeslices`). The
+    appropriate choice of operation will depend on the context and the quantity in
+    question.
+    """
+    from xarray.core.variable import Variable, _broadcast_compat_variables
+
+    if (isinstance(other, Variable)) and ("timeslice" in self.dims) != (
+        "timeslice" in getattr(other, "dims", [])
+    ):
+        raise ValueError(
+            "Broadcasting along the 'timeslice' dimension is required, but automatic "
+            "broadcasting is disabled. Please handle it explicitly using "
+            "`broadcast_timeslice` or `distribute_timeslice` (see `muse.timeslices`)."
+        )
+
+    # The rest of the function is copied directly from
+    # xarray.core.variable._broadcast_compat_data
+    if all(hasattr(other, attr) for attr in ["dims", "data", "shape", "encoding"]):
+        # `other` satisfies the necessary Variable API for broadcast_variables
+        new_self, new_other = _broadcast_compat_variables(self, other)
+        self_data = new_self.data
+        other_data = new_other.data
+        dims = new_self.dims
+    else:
+        # rely on numpy broadcasting rules
+        self_data = self.data
+        other_data = other
+        dims = self.dims
+    return self_data, other_data, dims
+
+
 if "__main__" == __name__:
-    run()
+    from unittest.mock import patch
+
+    with patch(
+        "xarray.core.variable._broadcast_compat_data", patched_broadcast_compat_data
+    ):
+        run()