Merge pull request #15451 from Fix broken 'differences from pandas' link
Fix broken 'differences from pandas' link
diff --git a/website/www/site/content/en/documentation/dsls/dataframes/overview.md b/website/www/site/content/en/documentation/dsls/dataframes/overview.md
index 294f925..e08c61b 100644
--- a/website/www/site/content/en/documentation/dsls/dataframes/overview.md
+++ b/website/www/site/content/en/documentation/dsls/dataframes/overview.md
@@ -45,7 +45,7 @@
In this example, the only traditional Beam type is the `Pipeline` instance. Otherwise the example is written completely with the DataFrame API. This is possible because the Beam DataFrame API includes its own IO operations (for example, `read_csv` and `to_csv`) based on the pandas native implementations. `read_*` and `to_*` operations support file patterns and any Beam-compatible file system. The grouping is accomplished with a group-by-key, and arbitrary pandas operations (in this case, `sum`) can be applied before the final write that occurs with `to_csv`.
-The Beam DataFrame API aims to be compatible with the native pandas implementation, with a few caveats detailed below in [Differences from standard pandas]({{< ref "#differences_from_standard_pandas" >}}).
+The Beam DataFrame API aims to be compatible with the native pandas implementation, with a few caveats detailed below in [Differences from standard pandas](/documentation/dsls/dataframes/differences-from-pandas/).
## Embedding DataFrames in a pipeline
