Implement using Config

Author: measrainsey

message_ix_models/tools/costs/config.py

@@ -20,6 +20,8 @@
 class Config:
     """Configuration for :mod:`.costs`."""
 
+    test_val: int = 2
+
     #: Base year for projections.
     base_year: int = BASE_YEAR
 

message_ix_models/tools/costs/demo.py

@@ -1,43 +1,88 @@
+from message_ix_models.tools.costs.config import Config
 from message_ix_models.tools.costs.projections import create_cost_projections
 
-# By default, the create_cost_projections() function will run for R12
+# Example 1: By default, the Config fill will run for:
+# R12
 # for the base suite of technologies,
 # with NAM as reference region,
 # using GDP as the cost driver,
-# and the updated data version.
+# and the updated data version
+# and outputs in MESSAGE format.
 # The function will also run for all SSP scenarios,
 # for all years from 2021 to 2100.
-inv, fix = create_cost_projections()
+default = Config()
+out_default = create_cost_projections(
+    node=default.node,
+    ref_region=default.ref_region,
+    base_year=default.base_year,
+    module=default.module,
+    method=default.method,
+    scenario_version=default.scenario_version,
+    scenario=default.scenario,
+    convergence_year=default.convergence_year,
+    fom_rate=default.fom_rate,
+    format=default.format,
+)
 
-# Example 1: Get cost projections for SSP2 scenario in R12,
+# Example 2: Get cost projections for SSP2 scenario in R12,
+# using WEU as the reference region,
+# with convergence as the method,
 # for materials technologies,
 # using GDP (updated data)
-inv2, fix2 = create_cost_projections(
-    sel_node="r12",
-    sel_ref_region="R12_NAM",
-    sel_base_year=2021,
-    sel_module="materials",
-    sel_scenario_version="updated",
-    sel_scenario="ssp2",
-    sel_method="gdp",
+# You can either put the inputs directly into the create_cost_projections function,
+# or you can create a Config object and pass that in.
+default = Config()
+
+# Option 1: Directly input the parameters
+out_materials_ssp2 = create_cost_projections(
+    node=default.node,
+    ref_region="R12_WEU",
+    base_year=default.base_year,
+    module="materials",
+    method="convergence",
+    scenario_version=default.scenario_version,
+    scenario="SSP2",
+    convergence_year=default.convergence_year,
+    fom_rate=default.fom_rate,
+    format=default.format,
 )
 
-# Example 2: Get cost projections in R11 (with WEU as reference region), using learning
-# (this will run for all SSP scenarios)
-inv, fix = create_cost_projections(
-    sel_node="r11",
-    sel_ref_region="R11_WEU",
-    sel_base_year=2021,
-    sel_method="learning",
-    sel_scenario_version="updated",
+# Option 2: Create a Config object and pass that in
+config = Config(
+    module="materials", scenario="SSP2", ref_region="R12_WEU", method="convergence"
 )
 
-# Example 3: Get cost projections in R12, using convergence
-inv, fix = create_cost_projections(
-    sel_node="r12",
-    sel_base_year=2021,
-    sel_method="convergence",
+out_materials_ssp2 = create_cost_projections(
+    node=config.node,
+    ref_region=config.ref_region,
+    base_year=config.base_year,
+    module=config.module,
+    method=config.method,
+    scenario_version=config.scenario_version,
+    scenario=config.scenario,
+    convergence_year=config.convergence_year,
+    fom_rate=config.fom_rate,
+    format=config.format,
 )
 
-# Example 4: Get cost projections in R11 using previous/original SSP scenarios
-inv, fix = create_cost_projections(sel_node="r11", sel_scenario_version="original")
+# Example 3: Get cost projections for SSP5 scenario in R12,
+# using LAM as the reference region,
+# with learning as the method,
+# for materials technologies,
+
+config = Config(
+    module="materials", scenario="SSP5", ref_region="R12_LAM", method="learning"
+)
+
+out_materials_ssp5 = create_cost_projections(
+    node=config.node,
+    ref_region=config.ref_region,
+    base_year=config.base_year,
+    module=config.module,
+    method=config.method,
+    scenario_version=config.scenario_version,
+    scenario=config.scenario,
+    convergence_year=config.convergence_year,
+    fom_rate=config.fom_rate,
+    format=config.format,
+)

message_ix_models/tools/costs/gdp.py

@@ -8,7 +8,7 @@
 
 
 # Function to read in (under-review) SSP data
-def process_raw_ssp_data(input_node, input_ref_region) -> pd.DataFrame:
+def process_raw_ssp_data(node, ref_region) -> pd.DataFrame:
     """Read in raw SSP data and process it
 
     This function takes in the raw SSP data (in IAMC format), aggregates \
@@ -38,22 +38,22 @@ def process_raw_ssp_data(input_node, input_ref_region) -> pd.DataFrame:
             (in units of billion US$2005/yr / million)
     """
     # Change node selection to upper case
-    node_up = input_node.upper()
+    node_up = node.upper()
 
     # Check if node selection is valid
     if node_up not in ["R11", "R12", "R20"]:
         print("Please select a valid region: R11, R12, or R20")
 
     # Set default reference region
-    if input_ref_region is None:
-        if input_node.upper() == "R11":
-            input_ref_region = "R11_NAM"
-        if input_node.upper() == "R12":
-            input_ref_region = "R12_NAM"
-        if input_node.upper() == "R20":
-            input_ref_region = "R20_NAM"
+    if ref_region is None:
+        if node.upper() == "R11":
+            ref_region = "R11_NAM"
+        if node.upper() == "R12":
+            ref_region = "R12_NAM"
+        if node.upper() == "R20":
+            ref_region = "R20_NAM"
     else:
-        input_ref_region = input_ref_region
+        ref_region = ref_region
 
     # Set data path for node file
     node_file = package_data_path("node", node_up + ".yaml")
@@ -162,7 +162,7 @@ def process_raw_ssp_data(input_node, input_ref_region) -> pd.DataFrame:
     )
 
     # If reference region is not in the list of regions, print error message
-    reference_region = input_ref_region.upper()
+    reference_region = ref_region.upper()
     if reference_region not in df.region.unique():
         print("Please select a valid reference region: " + str(df.region.unique()))
     # If reference region is in the list of regions, calculate GDP ratios
@@ -212,22 +212,22 @@ def process_raw_ssp_data(input_node, input_ref_region) -> pd.DataFrame:
 
 # Function to calculate adjusted region-differentiated cost ratios
 def calculate_indiv_adjusted_region_cost_ratios(
-    region_diff_df, input_node, input_ref_region, input_base_year
+    region_diff_df, node, ref_region, base_year
 ):
     df_gdp = (
-        process_raw_ssp_data(input_node=input_node, input_ref_region=input_ref_region)
+        process_raw_ssp_data(node=node, ref_region=ref_region)
         .query("year >= 2020")
         .drop(columns=["total_gdp", "total_population"])
     )
     df_cost_ratios = region_diff_df.copy()
 
     # If base year does not exist in GDP data, then use earliest year in GDP data
     # and give warning
-    base_year = int(input_base_year)
+    base_year = int(base_year)
     if int(base_year) not in df_gdp.year.unique():
         base_year = int(min(df_gdp.year.unique()))
         print(
-            f"Base year {input_base_year} not found in GDP data. \
+            f"Base year {base_year} not found in GDP data. \
                 Using {base_year} for GDP data instead."
         )
 
@@ -236,15 +236,15 @@ def calculate_indiv_adjusted_region_cost_ratios(
     # If specified node is R12, then use R12_NAM as the reference region
     # If specified node is R20, then use R20_NAM as the reference region
     # However, if a reference region is specified, then use that instead
-    if input_ref_region is None:
-        if input_node.upper() == "R11":
+    if ref_region is None:
+        if node.upper() == "R11":
             reference_region = "R11_NAM"
-        if input_node.upper() == "R12":
+        if node.upper() == "R12":
             reference_region = "R12_NAM"
-        if input_node.upper() == "R20":
+        if node.upper() == "R20":
             reference_region = "R20_NAM"
     else:
-        reference_region = input_ref_region
+        reference_region = ref_region
 
     gdp_base_year = df_gdp.query("year == @base_year").reindex(
         ["scenario_version", "scenario", "region", "gdp_ratio_reg_to_reference"], axis=1

message_ix_models/tools/costs/learning.py

@@ -11,7 +11,7 @@
 
 
 # Function to get GEA based cost reduction data
-def get_cost_reduction_data(input_module) -> pd.DataFrame:
+def get_cost_reduction_data(module) -> pd.DataFrame:
     """Get cost reduction data
 
     Raw data on cost reduction in 2100 for technologies are read from \
@@ -45,10 +45,10 @@ def get_cost_reduction_data(input_module) -> pd.DataFrame:
         .reset_index(drop=1)
     )
 
-    if input_module == "base":
+    if module == "base":
         return base_rates
 
-    elif input_module == "materials":
+    elif module == "materials":
         # Read in materials technology mapping file
         materials_file_path = package_data_path("costs", "technology_materials_map.csv")
         df_materials_tech = pd.read_csv(materials_file_path)
@@ -78,9 +78,7 @@ def get_cost_reduction_data(input_module) -> pd.DataFrame:
 
 
 # Function to get technology learning scenarios data
-def get_technology_learning_scenarios_data(
-    input_base_year, input_module
-) -> pd.DataFrame:
+def get_technology_learning_scenarios_data(base_year, module) -> pd.DataFrame:
     """Read in technology first year and learning scenarios data
 
     Raw data on technology first year and learning scenarios are read from \
@@ -90,7 +88,7 @@ def get_technology_learning_scenarios_data(
 
     Parameters
     ----------
-    input_base_year : int, optional
+    base_year : int, optional
         The base year, by default set to global BASE_YEAR
 
     Returns
@@ -109,9 +107,9 @@ def get_technology_learning_scenarios_data(
         pd.read_csv(file)
         .assign(
             first_technology_year=lambda x: np.where(
-                x.first_year_original > input_base_year,
+                x.first_year_original > base_year,
                 x.first_year_original,
-                input_base_year,
+                base_year,
             ),
         )
         .drop(columns=["first_year_original"])
@@ -122,10 +120,10 @@ def get_technology_learning_scenarios_data(
         )
     )
 
-    if input_module == "base":
+    if module == "base":
         return base_learn
 
-    elif input_module == "materials":
+    elif module == "materials":
         # Read in materials technology mapping file
         materials_file_path = package_data_path("costs", "technology_materials_map.csv")
         df_materials_tech = pd.read_csv(materials_file_path)
@@ -157,10 +155,10 @@ def get_technology_learning_scenarios_data(
 # Function to project reference region investment cost using learning rates
 def project_ref_region_inv_costs_using_learning_rates(
     regional_diff_df: pd.DataFrame,
-    input_node,
-    input_ref_region,
-    input_base_year,
-    input_module,
+    node,
+    ref_region,
+    base_year,
+    module,
 ) -> pd.DataFrame:
     """Project investment costs using learning rates for reference region
 
@@ -172,11 +170,11 @@ def project_ref_region_inv_costs_using_learning_rates(
     ----------
     regional_diff_df : pandas.DataFrame
         Dataframe output from :func:`get_weo_region_differentiated_costs`
-    input_node : str, optional
+    node : str, optional
         The reference node, by default "r12"
-    input_ref_region : str, optional
+    ref_region : str, optional
         The reference region, by default None (defaults set in function)
-    input_base_year : int, optional
+    base_year : int, optional
         The base year, by default set to global BASE_YEAR
 
     Returns
@@ -191,21 +189,21 @@ def project_ref_region_inv_costs_using_learning_rates(
     """
 
     # Set default reference region
-    if input_ref_region is None:
-        if input_node.upper() == "R11":
+    if ref_region is None:
+        if node.upper() == "R11":
             reference_region = "R11_NAM"
-        if input_node.upper() == "R12":
+        if node.upper() == "R12":
             reference_region = "R12_NAM"
-        if input_node.upper() == "R20":
+        if node.upper() == "R20":
             reference_region = "R20_NAM"
     else:
-        reference_region = input_ref_region
+        reference_region = ref_region
 
     # Get cost reduction data
-    df_cost_reduction = get_cost_reduction_data(input_module)
+    df_cost_reduction = get_cost_reduction_data(module)
 
     # Get learning rates data
-    df_learning = get_technology_learning_scenarios_data(input_base_year, input_module)
+    df_learning = get_technology_learning_scenarios_data(base_year, module)
 
     # Merge cost reduction data with learning rates data
     df_learning_reduction = df_learning.merge(
@@ -221,7 +219,7 @@ def project_ref_region_inv_costs_using_learning_rates(
             cost_region_2100=lambda x: x.reg_cost_base_year
             - (x.reg_cost_base_year * x.cost_reduction),
             b=lambda x: (1 - PRE_LAST_YEAR_RATE) * x.cost_region_2100,
-            r=lambda x: (1 / (LAST_MODEL_YEAR - input_base_year))
+            r=lambda x: (1 / (LAST_MODEL_YEAR - base_year))
             * np.log((x.cost_region_2100 - x.b) / (x.reg_cost_base_year - x.b)),
             reference_region=reference_region,
         )