Linting

Author: niekdejonge

ms2query/benchmarking/EvaluateMethods.py

@@ -61,8 +61,8 @@ def benchmark_exact_matching_within_ionmode(
 
         For each inchikey with more than 1 spectrum the spectra are split in two sets. Half for each inchikey is added
         to the library (training set), for the other half predictions are made. Thereby there is always an exact match
-        avaialable. Only the highest ranked prediction is considered correct if the correct inchikey is predicted. An accuracy per
-        inchikey is calculated followed by calculating the average.
+        avaialable. Only the highest ranked prediction is considered correct if the correct inchikey is predicted.
+        An accuracy per inchikey is calculated followed by calculating the average.
         """
         selected_spectra = subset_spectra_on_ionmode(self.validation_spectrum_set, ionmode)
 

ms2query/benchmarking/SpectrumDataSet.py

@@ -112,7 +112,8 @@ def subset_spectra(self, spectrum_indexes) -> "SpectraWithFingerprints":
         # Only keep the fingerprints for which we have inchikeys.
         # Important note: This is not a deep copy!
         # And the fingerprint is not reset (so it is not always actually matching the most common inchi)
-        new_instance.inchikey_fingerprint_pairs = {inchikey: self.inchikey_fingerprint_pairs[inchikey] for inchikey in new_instance.spectrum_indexes_per_inchikey.keys()}
+        new_instance.inchikey_fingerprint_pairs = {inchikey: self.inchikey_fingerprint_pairs[inchikey] for inchikey
+                                                   in new_instance.spectrum_indexes_per_inchikey.keys()}
         return new_instance
 
 

ms2query/benchmarking/reference_methods/PredictMS2DeepScoreSimilarity.py

@@ -21,7 +21,8 @@ def predict_top_ms2deepscores(
         k: Number of highest matches to return
 
     Returns:
-        List[List[int]: indexes of highest scores and the value for the highest score. Per query embedding the top k highest indexes are given.
+        List[List[int]: indexes of highest scores and the value for the highest score.
+        Per query embedding the top k highest indexes are given.
         List[List[float]: the highest scores.
     """
     top_indexes_per_batch = []

ms2query/benchmarking/reference_methods/predict_using_closest_tanimoto.py

@@ -25,9 +25,9 @@ def predict_using_closest_tanimoto(
     return inchikeys_of_best_match, highest_scores
 
 
-def predict_using_closest_tanimoto_single_spectrum(spectra_with_embeddings, single_spectrum_with_embeddings,
-                                                   nr_of_closest_inchikeys_to_select,
-                                                   nr_of_inchikeys_with_highest_ms2deepscore_to_select) -> Tuple[str, float]:
+def predict_using_closest_tanimoto_single_spectrum(
+        spectra_with_embeddings, single_spectrum_with_embeddings,
+        nr_of_closest_inchikeys_to_select, nr_of_inchikeys_with_highest_ms2deepscore_to_select) -> Tuple[str, float]:
     if len(single_spectrum_with_embeddings.spectra) != 1:
         raise ValueError("expected a single spectrum")
     ms2deepscores = cosine_similarity_matrix(single_spectrum_with_embeddings.embeddings,

ms2query/benchmarking/reference_methods/predict_with_integrated_similarity_flow.py

@@ -82,7 +82,8 @@ def integrated_similarity_flow(
     predicted_scores: List[float], similarities: np.ndarray, nr_of_spectra_per_inchikey: List[float]
 ) -> List[float]:
     """Compute the confidence of the prediction for each candidate.
-    Integrated similarity flow (ISF) scores are calculated using the similarity of candidates among each other and their distance to the query spectrum.
+    Integrated similarity flow (ISF) scores are calculated using the similarity of candidates among each other
+    and their distance to the query spectrum.
 
     Args:
         distances (list): Distances of the candidates to the query spectrum in the chemical space.