change over to dataframes, clean up types and pyproject

Author: karpdave

pyproject.toml

@@ -70,7 +70,7 @@ line-length = 120
 fix = true
 
 [tool.ruff.lint]
-exclude = ["**/*.py", "**/*.pyi"] # TEMPORARY!!!!
+exclude = ["**/capex_project.py", "**/model.py", "**/npa_project.py", "**/params.py", "**/utils.py", "**/web_params.py", "skeleton.pyi"] # TEMPORARY!!!!
 select = [
     # flake8-2020
     "YTT",

src/npa_howtopay/capex_project.py

@@ -1,28 +1,52 @@
-class CapexProject:
-    project_year: int
-    original_cost: float
-    depreciation_lifetime: int
-    # depreciation_schedule: str # "straight line" or "accelerated"
+from attrs import define
+from params import InputParams
 
-    @property
-    def get_depreciation(self, year: int) -> float:
-        if self.project_year < year <= self.project_year + self.depreciation_lifetime:
-            return self.original_cost / self.depreciation_lifetime
-        else:
-            return 0
+import polars as pl
 
+## All dataframes used by functions in this class will have the following columns:
+# project_year: int
+# original_cost: float
+# depreciation_lifetime: int
+## (someday)depreciation_schedule: str # "straight line" or "accelerated"
 
-def non_lpp_gas_capex_projects(year: int, input_params: InputParams) -> list[CapexProject]:
+
+# functions for generating dataframe rows for capex projects
+def get_synthetic_initial_capex_projects(
+    start_year: int, initial_ratebase: float, depreciation_lifetime: int
+) -> pl.DataFrame:
+    total_weight = (depreciation_lifetime * (depreciation_lifetime + 1) / 2) / depreciation_lifetime
+    est_original_cost_per_year = initial_ratebase / total_weight
+    return pl.DataFrame({
+        "project_year": range(start_year, start_year + depreciation_lifetime),
+        "original_cost": est_original_cost_per_year,
+        "depreciation_lifetime": depreciation_lifetime,
+    })
+
+
+def get_non_lpp_gas_capex_projects(year: int, input_params: InputParams) -> pl.DataFrame:
+    pass
+
+
+def get_lpp_gas_capex_projects(year: int, input_params: InputParams, npas_this_year: NpaProject) -> pl.DataFrame:
+    pass
+
+
+def get_non_npa_electric_capex_projects(year: int, input_params: InputParams) -> pl.DataFrame:
+    pass
+
+
+def get_grid_upgrade_capex_projects(year: int, input_params: InputParams, npas_projects: pl.DataFrame) -> pl.DataFrame:
     pass
 
 
-def lpp_gas_capex_projects(year: int, input_params: InputParams, npas_this_year: NpaProject) -> list[CapexProject]:
+def get_npa_capex_projects(year: int, input_params: InputParams, npas_projects: pl.DataFrame) -> pl.DataFrame:
     pass
 
 
-def non_npa_electric_capex_projects(year: int, input_params: InputParams) -> list[CapexProject]:
+# functions for computing things given a dataframe of capex projects
+def compute_ratebase_from_capex_projects(df: pl.DataFrame, year: int) -> float:
     pass
 
 
-def grid_upgrade_capex_projects(year: int, input_params: InputParams, npas_this_year: NpaProject) -> list[CapexProject]:
+def compute_depreciation_expense_from_capex_projects(df: pl.DataFrame, year: int) -> float:
     pass

src/npa_howtopay/model.py

@@ -16,48 +16,46 @@ def compute_blue_columns(year: int, gas_params: GasParams, shared_params: Shared
 
 
 def run_model(scenario_params: ScenarioParams, input_params: InputParams, npa_projects: pl.DataFrame) -> pl.DataFrame:
-    # initialize all the state
-    gas_ratebase = 0
-    electric_ratebase = 0
-    num_gas_users = input_params.num_users
-    num_electric_users = input_params.num_users
-
-    gas_capex_projects = []
-    electric_capex_projects = []
+    gas_capex_projects = pl.DataFrame()
+    electric_capex_projects = pl.DataFrame()
 
     output_df = pl.DataFrame()
 
+    live_params = input_params
+
     for year in range(scenario_params.start_year, scenario_params.end_year):
         # INFLATION - probably looks like evolving input_params to current_params
-        do_cost_inflation(year, input_params)
+        live_params = do_cost_inflation(year, live_params)
 
         # get the npas for this year
-        npas_this_year = get_info_about_projects(npa_projects, year)  # polars groupby year, weighted_sum
+        npas_this_year = npa_projects.filter(pl.col("year") == year)
 
         ####### ticket
         # gas capex
-        gas_capex_projects.append(
+        gas_capex_projects = gas_capex_projects.vstack(
             non_lpp_gas_capex_projects(year, input_params)
         )  # maybe has different depreciation schedule
-        gas_capex_projects.append(lpp_gas_capex_projects(year, input_params, npas_this_year))
+        gas_capex_projects = gas_capex_projects.vstack(lpp_gas_capex_projects(year, input_params, npas_this_year))
 
         ####### ticket
         # electric capex
-        electric_capex_projects.append(non_npa_electric_capex_projects(year, input_params))
-        electric_capex_projects.append(grid_upgrade_capex_projects(year, input_params, npa_projects))
+        electric_capex_projects = electric_capex_projects.vstack(non_npa_electric_capex_projects(year, input_params))
+        electric_capex_projects = electric_capex_projects.vstack(
+            grid_upgrade_capex_projects(year, input_params, npa_projects)
+        )
 
         # npa capex
         if scenario_params.capex_opex == "capex":
-            npa_capex = get_npa_capex_project(npas_this_year)
+            npa_capex = get_npa_capex_projects(npas_this_year, year, input_params)
             if scenario_params.gas_electric == "gas":
-                gas_capex_projects.append(npa_capex)
+                gas_capex_projects = gas_capex_projects.vstack(npa_capex)
             elif scenario_params.gas_electric == "electric":
-                electric_capex_projects.append(npa_capex)
+                electric_capex_projects = electric_capex_projects.vstack(npa_capex)
 
-        gas_ratebase = update_ratebase(year, gas_ratebase, gas_capex_projects)
-        electric_ratebase = update_ratebase(year, electric_ratebase, electric_capex_projects)
+        gas_ratebase = get_ratebase_from_capex_projects(year, gas_capex_projects)
+        electric_ratebase = get_ratebase_from_capex_projects(year, electric_capex_projects)
 
-        num_gas_users -= npas_this_year.num_converts
+        ## TODO: opex stuff?
 
         ####### ticket
         intermediate_columns = compute_blue_columns()  # returns pl.Series

src/npa_howtopay/npa_project.py

@@ -1,10 +1,26 @@
 from attrs import define
 
+import polars as pl
 
-@define
-class NpaProject:
-    year: int
-    num_converts: int
-    pipe_value_per_user: float
-    pipe_decomm_cost_per_user: float
-    # npa_install_costs: float # maybe? is this in 2025 dollars?
+
+## All dataframes used by functions in this class will have the following columns:
+# year: int
+# num_converts: int
+# pipe_value_per_user: float
+# pipe_decomm_cost_per_user: float
+
+# note: could represent scattershot electrification as a row with pipe_value_per_user and pipe_decomm_cost_per_user set to 0
+
+
+def compute_num_converts_from_df(df: pl.DataFrame, year: int) -> int:
+    return df.filter(pl.col("year") <= year).sum("num_converts")
+
+
+def compute_avoided_pipe_costs_from_df(df: pl.DataFrame, year: int) -> float:
+    # wrong, needs to multiply by num_converts
+    # return df.filter(pl.col("year") == year).sum("pipe_value_per_user")
+    pass
+
+
+def compute_pipe_decomm_cost_from_df(df: pl.DataFrame) -> pl.DataFrame:
+    pass