From ee00432cd72fa55ddb3e67c8c7d7b72b2520ccd1 Mon Sep 17 00:00:00 2001
From: Robbe Sneyders <robbe.sneyders@ml6.eu>
Date: Thu, 15 Feb 2024 09:53:14 +0100
Subject: [PATCH] Add docs

---
 docs/components/components.md | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)

diff --git a/docs/components/components.md b/docs/components/components.md
index 914c75f3..dfa8c231 100644
--- a/docs/components/components.md
+++ b/docs/components/components.md
@@ -1,3 +1,5 @@
+from distributed import Client
+
 # Components
 
 Fondant makes it easy to build data preparation pipelines leveraging reusable components. Fondant
@@ -65,6 +67,32 @@ this data can be accessed using `dataframe["image"]`.
 The `transform` method should return a single dataframe, with the columns complying to the
 schema defined by the `produces` section of the component specification.
 
+### Configuring Dask
+
+You can configure the Dask client based on the needs of your component by overriding the 
+`dask_client` method:
+
+```python
+import os
+
+from dask.distributed import Client, LocalCluster
+from fondant.component import PandasTransformComponent
+
+class Component(PandasTransformComponent):
+  
+    def dask_client(self) -> Client:
+        """Initialize the dask client to use for this component."""
+        cluster = LocalCluster(
+            processes=True,
+            n_workers=os.cpu_count(),
+            threads_per_worker=1,
+        )
+        return Client(cluster)
+```
+
+The default Dask client is configured to work with processes, the same amount of workers as 
+logical CPUs available, and on thread per worker.
+
 ## Component types
 
 We can distinguish two different types of components: