More linting. Ignore ESP file. Clean redundancy in CI.

.github/workflows/eluc.yml

@@ -30,7 +30,7 @@ jobs:
         pip install pylint
         if [ -f requirements.txt ]; then pip install -r requirements.txt; fi
     - name: Lint with PyLint
-      run: pylint --ignore="demo" --recursive=y --fail-under=9 ./*
+      run: pylint ./*
     - name: Run unit tests
       run: python -m unittest
 

use_cases/eluc/.pylintrc

@@ -1,10 +1,14 @@
 [MASTER]
-ignore=demo
+ignore=demo, prescriptors/esp
+
+recursive=y
+
+fail-under=9.0
 
 jobs=0
 
 max-line-length=120
 
 suggestion-mode=yes
 
-good-names=X_train, X_val, X_test, y_train, y_val, y_test, X, Y, y, X_test_scaled
+good-names=X_train, X_val, X_test, y_train, y_val, y_test, X, Y, y, X_test_scaled, df, da, ds, i, n

use_cases/eluc/README.md

@@ -23,7 +23,7 @@ BLUE simulations with committed emissions could be used to estimate the long-ter
 "Committed emissions" means all the emissions that are caused by a land-use change event are attributed to the year
 of the event.
 BLUE (bookkeeping of land use emissions) is a bookkeeping model that attributes carbon fluxes to land use activities.
-See [BLUE: Bookkeeping of land use emissions](https://doi.org/10.1002/2014GB004997) for more details. 
+See [BLUE: Bookkeeping of Land Use Emissions](https://doi.org/10.1002/2014GB004997) for more details.
 
 ### LUC
 

use_cases/eluc/data/conversion.py

@@ -58,6 +58,8 @@ def construct_countries_df():
     for i in range(len(countries_df)):
         old_abbrev = countries_df.iloc[i]["abbrevs"]
         if old_abbrev in MANUAL_MAP and MANUAL_MAP[old_abbrev] in codes_df["Numeric code"].unique():
-            countries_df.iloc[i]["abbrevs"] = codes_df[codes_df["Numeric code"] == MANUAL_MAP[old_abbrev]]["Alpha-2 code"].iloc[0]
+            numeric_code = codes_df["Numeric code"]
+            old_abbrev = MANUAL_MAP[old_abbrev]
+            countries_df.iloc[i]["abbrevs"] = codes_df[numeric_code == old_abbrev]["Alpha-2 code"].iloc[0]
 
     return countries_df

use_cases/eluc/prescriptors/nsga2/candidate.py

@@ -43,7 +43,7 @@ def __init__(self, in_size: int, hidden_size: int, out_size: int,
         self.parents = parents
 
     @classmethod
-    def from_crossover(cls, parent1, parent2, p_mutation: float, gen: int, cand_id: int):
+    def from_crossover(cls, parent1, parent2, p_mutation: float, gen: int, cand_id: int) -> "Candidate":
         """
         Crossover two parents to create a child.
         Take a random 50/50 choice of either parent's weights

use_cases/eluc/prescriptors/nsga2/nsga2_utils.py

@@ -64,7 +64,7 @@ def calculate_crowding_distance(front):
     n_objectives = len(front[0].metrics)
     distances = [0 for _ in range(len(front))]
     for m in range(n_objectives):
-        front.sort(key=lambda c: c.metrics[m])
+        front.sort(key=lambda candidate: candidate.metrics[m])
         obj_min = front[0].metrics[m]
         obj_max = front[-1].metrics[m]
         distances[0] = float('inf')

use_cases/eluc/prescriptors/nsga2/torch_prescriptor.py

@@ -32,7 +32,7 @@ def __init__(self,
                  batch_size: int,
                  candidate_params: dict,
                  seed_dir=None):
-        
+
         self.candidate_params = candidate_params
         self.pop_size = pop_size
         self.n_generations = n_generations
@@ -149,7 +149,7 @@ def _select_parents(self, candidates: list[Candidate], n_parents: int) -> list[C
             for candidate, distance in zip(front, crowding_distance):
                 candidate.distance = distance
             if len(parents) + len(front) > n_parents:  # If adding this front exceeds num_parents
-                front = sorted(front, key=lambda c: c.distance, reverse=True)
+                front = sorted(front, key=lambda candidate: candidate.distance, reverse=True)
                 parents += front[:n_parents - len(parents)]
                 break
             parents += front
@@ -169,7 +169,7 @@ def _make_new_pop(self, parents: list[Candidate], pop_size: int, gen:int) -> lis
         Makes new population by creating children from parents.
         We use tournament selection to select parents for crossover.
         """
-        sorted_parents = sorted(parents, key=lambda c: (c.rank, -c.distance))
+        sorted_parents = sorted(parents, key=lambda candidate: (candidate.rank, -candidate.distance))
         children = []
         for i in range(pop_size):
             parent1, parent2 = self._tournament_selection(sorted_parents)
@@ -226,23 +226,23 @@ def _record_gen_results(self, gen: int, candidates: list[Candidate], save_path:
         Save the pareto front to disk.
         """
         # Save statistics of candidates
-        gen_results = [c.record_state() for c in candidates]
+        gen_results = [candidate.record_state() for candidate in candidates]
         gen_results_df = pd.DataFrame(gen_results)
         gen_results_df.to_csv(save_path / f"{gen}.csv", index=False)
 
         # Save rank 1 candidate state dicts
         (save_path / f"{gen}").mkdir(parents=True, exist_ok=True)
-        pareto_candidates = [c for c in candidates if c.rank == 1]
-        for c in pareto_candidates:
-            torch.save(c.state_dict(), save_path / f"{gen}" / f"{c.gen}_{c.cand_id}.pt")
+        pareto_candidates = [candidate for candidate in candidates if candidate.rank == 1]
+        for candidate in pareto_candidates:
+            torch.save(candidate.state_dict(), save_path / f"{gen}" / f"{candidate.gen}_{candidate.cand_id}.pt")
 
     def _record_candidate_avgs(self, gen: int, candidates: list[Candidate]) -> dict:
         """
         Gets the average eluc and change for a population of candidates.
         """
         avg_eluc = np.mean([c.metrics[0] for c in candidates])
         avg_change = np.mean([c.metrics[1] for c in candidates])
-        return {"gen": gen, "eluc": avg_eluc, "change": avg_change}     
+        return {"gen": gen, "eluc": avg_eluc, "change": avg_change}
 
     def prescribe_land_use(self, context_df: pd.DataFrame, **kwargs) -> pd.DataFrame:
         """

use_cases/eluc/requirements.txt

@@ -171,6 +171,7 @@ traitlets==5.9.0
 transformers==4.31.0
 typing_extensions==4.6.2
 uri-template==1.2.0
+wcwidth==0.2.13
 webcolors==1.13
 webencodings==0.5.1
 websocket-client==1.5.3

use_cases/eluc/tests/test_nsga2.py

@@ -21,7 +21,7 @@ def get_fields(df):
     Dummy fields method to create the encoder for dummy data.
     """
     fields_df = df[constants.CAO_MAPPING["context"] + constants.CAO_MAPPING["actions"] + ["ELUC"]].astype("float64")
-    fields = dict()
+    fields = {}
     for col in constants.CAO_MAPPING["context"] + constants.CAO_MAPPING["actions"] + ["ELUC"]:
         # Set range of land and diff land uses manually to their true ranges because they
         # do not need to be scaled
@@ -52,7 +52,7 @@ def get_fields(df):
         "valued": "CONTINUOUS"
     }
 
-    return fields 
+    return fields
 
 class TestTorchPrescriptor(unittest.TestCase):
     """