Add large_grow example

examples/atomic_dict/large_grow.py

@@ -0,0 +1,39 @@
+import threading
+import time
+
+import numpy as np
+from cereggii import AtomicDict
+
+
+keys_count = 10_000_000
+keys = np.random.default_rng(seed=0).choice(2**5 * keys_count, size=keys_count, replace=False)  # distinct
+keys = keys.tolist()
+
+
+def large_grow(dict_factory, num_threads):
+    for n in num_threads:
+        d = dict_factory()
+
+        def thread(i, keys):
+            for _ in range(i, keys_count, n):
+                d[keys[_]] = None
+
+        threads = []
+        for i in range(n):
+            t = threading.Thread(target=thread, args=(i, keys.copy()))
+            threads.append(t)
+
+        started = time.time()
+        for t in threads:
+            t.start()
+        for t in threads:
+            t.join()
+
+        took = time.time() - started
+        print(f" - Took {took:2.3f}s with {n} threads.")
+
+
+print(f"\nInserting {keys_count} keys using dict:")
+large_grow(dict, [1, 3])
+print(f"\nInserting {keys_count} keys using AtomicDict:")
+large_grow(AtomicDict, [1, 3])