Use isinstance instead of type

ranx/meta/compare.py

@@ -65,7 +65,7 @@ def compare(
         Report: See report.
     """
     metrics = format_metrics(metrics)
-    assert all(type(m) == str for m in metrics), "Metrics error"
+    assert all(isinstance(m, str) for m in metrics), "Metrics error"
 
     model_names = []
     results = defaultdict(dict)

ranx/meta/evaluate.py

@@ -11,7 +11,7 @@
 
 
 def format_metrics(metrics: Union[List[str], str]) -> List[str]:
-    if type(metrics) == str:
+    if isinstance(metrics, str):
         metrics = [metrics]
     return metrics
 
@@ -43,15 +43,15 @@ def extract_metric_and_params(metric):
 def convert_qrels(qrels):
     if type(qrels) == Qrels:
         return qrels.to_typed_list()
-    elif type(qrels) == dict:
+    elif isinstance(qrels, dict):
         return python_dict_to_typed_list(qrels, sort=True)
     return qrels
 
 
 def convert_run(run):
     if type(run) == Run:
         return run.to_typed_list()
-    elif type(run) == dict:
+    elif isinstance(run, dict):
         return python_dict_to_typed_list(run, sort=True)
     return run
 
@@ -133,7 +133,7 @@ def evaluate(
     _qrels = convert_qrels(qrels)
     _run = convert_run(run)
     metrics = format_metrics(metrics)
-    assert all(type(m) == str for m in metrics), "Metrics error"
+    assert all(isinstance(m, str) for m in metrics), "Metrics error"
 
     # Compute metrics ----------------------------------------------------------
     metric_scores_dict = {}

ranx/meta/plot.py

@@ -25,7 +25,7 @@ def plot(
 
     _qrels = qrels.to_typed_list()
 
-    if type(runs) == list:
+    if isinstance(runs, list):
         _runs = [run.to_typed_list() for run in runs]
         names = [
             run.name if run.name is not None else f"run_{i+1}"