linting

Author: ncfrey

.github/workflows/push.yml

@@ -32,6 +32,7 @@ jobs:
           python-version: ${{ matrix.python }}
       - run: "python -m pip install -r requirements.in"
       - run: "python -m pip install -r requirements-dev.in"
+      - run: "python -m pip install -r requirements-mgm.in"
       - run: "python -m pip install --editable ."
       - run: "python -m pytest"
       - env:

src/lobster/_imports.py

@@ -1 +0,0 @@
-from lightning_utilities.core.imports import RequirementCache

src/lobster/data/_calm_datamodule.py

@@ -4,13 +4,13 @@
 from typing import Any, Callable, Iterable, Optional, Sequence, TypeVar, Union
 
 import torch.utils.data
-from lobster.transforms import Transform
 from lightning import LightningDataModule
 from torch import Generator
 from torch.utils.data import DataLoader, Sampler
 
 from lobster.datasets._calm_dataset import CalmDataset
 from lobster.tokenization import PmlmTokenizerTransform
+from lobster.transforms import Transform
 
 T = TypeVar("T")
 

src/lobster/data/_cmap_datamodule.py

@@ -2,12 +2,13 @@
 from typing import Any, Callable, Iterable, Optional, Sequence, TypeVar, Union
 
 import torch
-from lobster.transforms import Transform
 from lightning import LightningDataModule
 from pandas import DataFrame
 from torch import Generator, Tensor
 from torch.utils.data import DataLoader, Dataset, Sampler
 
+from lobster.transforms import Transform
+
 T = TypeVar("T")
 
 
@@ -247,9 +248,7 @@ def __getitem__(self, index: int) -> T:
         if self._target_transform_fn is not None:
             y = self._target_transform_fn(y)
 
-        if len(self._target_columns) > 1 and not all(
-            isinstance(y_val, Tensor) for y_val in y
-        ):
+        if len(self._target_columns) > 1 and not all(isinstance(y_val, Tensor) for y_val in y):
             y = tuple(Tensor(y_val) for y_val in y)
 
         elif not isinstance(y, Tensor):

src/lobster/data/_dataframe_dataset_in_memory.py

@@ -1,9 +1,10 @@
-import torch
-import edlib
 import heapq
-import numpy as np
 from typing import Optional
 
+import edlib
+import numpy as np
+import torch
+
 
 class FarthestFirstTraversal:
     def __init__(
@@ -53,10 +54,7 @@ def str_fft(self, inputs: list[str]):
         inputs = [inputs[i] for i in perm]
         centroids = [inputs[i] for i in range(self._k)]
         while len(centroids) < self._num_samples:
-            dist = [
-                min(self._levenshtein(str1, str2) for str2 in centroids)
-                for str1 in inputs
-            ]
+            dist = [min(self._levenshtein(str1, str2) for str2 in centroids) for str1 in inputs]
             farthest = dist.index(max(dist))
             if inputs[farthest] in centroids:
                 break
@@ -82,12 +80,14 @@ def _levenshtein(s1, s2):
 
         return previous_row[-1]
 
+
 def edit_dist(x: str, y: str):
     """
     Computes the edit distance between two strings.
     """
     return edlib.align(x, y)["editDistance"]
 
+
 def ranked_fft(
     library: np.ndarray,
     ranking_scores: Optional[np.ndarray] = None,
@@ -138,10 +138,7 @@ def ranked_fft(
             neg_dist, score, idx, num_checked = heapq.heappop(pq)
             # Check if the top of the heap has been checked against all currently selected sequences
             if num_checked < len(selected):
-                min_dist = min(
-                    edit_dist(library[idx], library[selected[i]])
-                    for i in range(num_checked, len(selected))
-                )
+                min_dist = min(edit_dist(library[idx], library[selected[i]]) for i in range(num_checked, len(selected)))
                 min_dist = min(min_dist, -neg_dist)
                 heapq.heappush(pq, (-min_dist, score, idx, len(selected)))
             else: