Improve prune_at parameter description for `gensim.corpora.Dictiona…

…ry` (#2128)

* Make clear `prune_at` documentation

According to the code, the `prune_at` parameter in `Dictionary.__init__` and `add_documents` is only for reducing memory usage, and has no guarantee on correctness, but the documentation of this parameter was confusing to users.

* add link to method

gensim/corpora/dictionary.py

@@ -56,7 +56,9 @@ def __init__(self, documents=None, prune_at=2000000):
         documents : iterable of iterable of str, optional
             Documents to be used to initialize the mapping and collect corpus statistics.
         prune_at : int, optional
-            Dictionary will keep no more than `prune_at` words in its mapping, to limit its RAM footprint.
+            Dictionary will try to keep no more than `prune_at` words in its mapping, to limit its RAM
+            footprint, the correctness is not guaranteed.
+            Use :meth:`~gensim.corpora.dictionary.Dictionary.filter_extremes` to perform proper filtering.
 
         Examples
         --------
@@ -172,7 +174,9 @@ def add_documents(self, documents, prune_at=2000000):
         documents : iterable of iterable of str
             Input corpus. All tokens should be already **tokenized and normalized**.
         prune_at : int, optional
-            Dictionary will keep no more than `prune_at` words in its mapping, to limit its RAM footprint.
+            Dictionary will try to keep no more than `prune_at` words in its mapping, to limit its RAM
+            footprint, the correctness is not guaranteed.
+            Use :meth:`~gensim.corpora.dictionary.Dictionary.filter_extremes` to perform proper filtering.
 
         Examples
         --------