Adapt examples, changes to flow naming and df export

examples/example1.ipynb

@@ -103,7 +103,9 @@
    "metadata": {},
    "source": [
     "## 4. Define the MFA System equations\n",
-    "We define a class with our system equations in the compute method. Afterwards we create an instance of this class, using the input data defined above. The class (system equations) can then easily be reused with different input data."
+    "We define a class with our system equations in the compute method. Afterwards we create an instance of this class, using the input data defined above. The class (system equations) can then easily be reused with different input data.\n",
+    "\n",
+    "We just need to define the compute method with our system equations, as all the other things we need are inherited from the MFASystem class."
    ]
   },
   {
@@ -113,8 +115,7 @@
    "outputs": [],
    "source": [
     "class SimpleMFA(MFASystem):\n",
-    "    \"\"\"We just need to define the compute method with our system equations,\n",
-    "    as all the other things we need are inherited from the MFASystem class.\"\"\"\n",
+    "\n",
     "    def compute(self):\n",
     "        self.flows['sysenv => process 1'][...] = self.parameters['D']  # the elipsis slice [...] ensures the dimensionality of the flow is not changed\n",
     "        self.flows['process 1 => process 2'][...] = 1 / (1 - self.parameters['alpha']) * self.parameters['D']\n",

pyproject.toml

@@ -18,6 +18,9 @@ plotly = "^5.18.0"
 pydantic = "^2.8.2"
 ruff = "^0.5.6"
 docformatter = "^1.7.5"
+openpyxl = "^3.1.5"
+nbformat = "^5.10.4"
+jupyter = "^1.1.1"
 
 [tool.poetry.group.test]
 optional = true
@@ -35,6 +38,10 @@ autodoc-pydantic = "^2.2.0"
 sphinx-rtd-theme = "^2.0.0"
 myst-parser = "^4.0.0"
 
+
+[tool.poetry.group.dev.dependencies]
+ipykernel = "^6.29.5"
+
 [build-system]
 requires = ["poetry-core"]
 build-backend = "poetry.core.masonry.api"

sodym/export/data_writer.py

@@ -8,6 +8,7 @@
 from pydantic import BaseModel as PydanticBaseModel
 
 from ..mfa_system import MFASystem
+from .helper import to_valid_file_name
 
 
 class DataWriter(PydanticBaseModel):
@@ -28,16 +29,16 @@ def export_mfa_flows_to_csv(self, mfa: MFASystem, export_directory: str):
         if not os.path.exists(export_directory):
             os.makedirs(export_directory)
         for flow_name, flow in mfa.flows.items():
-            path_out = os.path.join(export_directory, f'{flow_name.replace(" => ", "__2__")}.csv')
-            flow.to_df().to_csv(path_out, index=False)
+            path_out = os.path.join(export_directory, f'{to_valid_file_name(flow_name)}.csv')
+            flow.to_df().to_csv(path_out)
         logging.info(f'Data saved in directory {export_directory}')
 
     def export_mfa_stocks_to_csv(self, mfa: MFASystem, export_directory: str):
         if not os.path.exists(export_directory):
             os.makedirs(export_directory)
         for stock_name, stock in mfa.stocks.items():
             df = stock.stock.to_df()
-            path_out = os.path.join(export_directory, f'{stock_name}_stock.csv')
+            path_out = os.path.join(export_directory, f'{to_valid_file_name(stock_name)}_stock.csv')
             df.to_csv(path_out, index=False)
         logging.info(f'Data saved in directory {export_directory}')
 

sodym/export/helper.py

@@ -1,6 +1,8 @@
 from matplotlib import pyplot as plt
 import numpy as np
 from pydantic import BaseModel as PydanticBaseModel
+import re
+import unicodedata
 
 from ..named_dim_arrays import NamedDimArray
 from ..dimensions import DimensionSet
@@ -74,7 +76,7 @@ def fill_fig_ax(self):
             n_subplots = self.array.dims[self.subplot_dim].len
             nx = int(np.ceil(np.sqrt(n_subplots)))
             ny = int(np.ceil(n_subplots / nx))
-        self.fig_ax = plt.subplots(nx, ny, figsize=(10, 9))
+        self.fig_ax = plt.subplots(nx, ny, figsize=(10, 9), squeeze=False)
 
     def get_x_array_like_value_array(self):
         if self.x_array is None:
@@ -134,3 +136,16 @@ def list_of_slices(array: NamedDimArray, dim_name_to_slice) -> tuple[list[NamedD
         list_array = [array]
         list_name = [None]
     return list_array, list_name
+
+
+def to_valid_file_name(value: str) -> str:
+    """
+    Taken from https://github.com/django/django/blob/master/django/utils/text.py
+    Convert to ASCII. Convert spaces or repeated dashes to single dashes.
+    Remove characters that aren't alphanumerics, underscores, or hyphens.
+    Convert to lowercase. Also strip leading and trailing whitespace, dashes, and underscores.
+    """
+    value = str(value)
+    value = unicodedata.normalize('NFKD', value).encode('ascii', 'ignore').decode('ascii')
+    value = re.sub(r'[^\w\s-]', '', value.lower())
+    return re.sub(r'[-\s]', '_', value).strip('-_')

sodym/flow_helper.py

@@ -1,11 +1,14 @@
+from typing import Callable
+
 from .named_dim_arrays import Process, Flow
 from .dimensions import DimensionSet
 from .mfa_definition import FlowDefinition
+from .flow_naming import process_names_with_arrow
 
 
 def make_empty_flows(
         processes: dict[str, Process], flow_definitions: list[FlowDefinition],
-        dims: DimensionSet,
+        dims: DimensionSet, naming: Callable[[Process, Process], str] = process_names_with_arrow
         ) -> dict[str, Flow]:
     """Initialize all defined flows with zero values."""
     flows = {}
@@ -15,7 +18,11 @@ def make_empty_flows(
             to_process = processes[flow_definition.to_process_name]
         except KeyError:
             raise KeyError(f"Missing process required by flow definition {flow_definition}.")
+        if flow_definition.name_override is not None:
+            name = flow_definition.name_override
+        else:
+            name = naming(from_process, to_process)
         dim_subset = dims.get_subset(flow_definition.dim_letters)
-        flow = Flow(from_process=from_process, to_process=to_process, dims=dim_subset)
+        flow = Flow(from_process=from_process, to_process=to_process, name=name, dims=dim_subset)
         flows[flow.name] = flow
     return flows

sodym/flow_naming.py

@@ -0,0 +1,13 @@
+from .named_dim_arrays import Process
+
+
+def process_names_with_arrow(from_process: Process, to_process: Process) -> str:
+    return f"{from_process.name} => {to_process.name}"
+
+
+def process_names_no_spaces(from_process: Process, to_process: Process) -> str:
+    return f"{from_process.name}_to_{to_process.name}"
+
+
+def process_ids(from_process: Process, to_process: Process) -> str:
+    return f"F{from_process.id}_{to_process.id}"

sodym/mfa_definition.py

@@ -53,6 +53,7 @@ class FlowDefinition(DefinitionWithDimLetters):
 
     from_process_name: str = Field(validation_alias=AliasChoices("from_process_name", "from_process"))
     to_process_name: str = Field(validation_alias=AliasChoices("to_process_name", "to_process"))
+    name_override: Optional[str] = None
 
 
 class StockDefinition(DefinitionWithDimLetters):

sodym/named_dim_arrays.py

@@ -82,6 +82,11 @@ def sub_array_handler(self, definition):
     def shape(self):
         return self.dims.shape()
 
+    def set_values(self, values: np.ndarray):
+        assert isinstance(values, np.ndarray), "Values must be a numpy array."
+        assert values.shape == self.shape, "Values must have the same shape as the DimensionSet."
+        self.values = values
+
     def sum_values(self):
         return np.sum(self.values)
 
@@ -219,10 +224,12 @@ def __setitem__(self, keys, item):
         slice_obj = self.sub_array_handler(keys)
         slice_obj.values_pointer[...] = item.sum_values_to(slice_obj.dim_letters)
 
-    def to_df(self):
-        index = pd.MultiIndex.from_product([d.items for d in self.dims], names=self.dims.names)
-        df = index.to_frame(index=False)
-        df["value"] = self.values.flatten()
+    def to_df(self, index: bool = True):
+        multiindex = pd.MultiIndex.from_product([d.items for d in self.dims], names=self.dims.names)
+        df = pd.DataFrame({'value': self.values.flatten()})
+        df = df.set_index(multiindex)
+        if not index:
+            df = df.reset_index()
         return df
 
 
@@ -404,11 +411,6 @@ class Flow(NamedDimArray):
     from_process: Process
     to_process: Process
 
-    @model_validator(mode="after")
-    def flow_name_related_to_proccesses(self):
-        self.name = f"{self.from_process.name} => {self.to_process.name}"
-        return self
-
     @property
     def from_process_id(self):
         return self.from_process.id