From 23994e730f5460d6a6e3dd642dce1605ae270153 Mon Sep 17 00:00:00 2001
From: Miha Zgubic <miha.zgubic@invenialabs.co.uk>
Date: Mon, 16 May 2022 16:24:57 +0100
Subject: [PATCH] mention fitting transforms in the docs

---
 docs/src/examples.md   |  2 ++
 docs/src/transforms.md | 15 +++++++++++++--
 2 files changed, 15 insertions(+), 2 deletions(-)

diff --git a/docs/src/examples.md b/docs/src/examples.md
index 0ca7392..d2c307c 100644
--- a/docs/src/examples.md
+++ b/docs/src/examples.md
@@ -7,6 +7,8 @@ First we load some hourly weather data:
 ```jldoctest example
 julia> using DataFrames, Dates, FeatureTransforms
 
+julia> using FeatureTransforms: fit!
+
 julia> df = DataFrame(
             :time => DateTime(2018, 9, 10):Hour(1):DateTime(2018, 9, 10, 23),
             :temperature => [10.6, 9.5, 8.9, 8.9, 8.4, 8.4, 7.7, 8.9, 11.7, 13.9, 16.2, 17.7, 18.9, 20.0, 21.2, 21.7, 21.7, 21.2, 20.0, 18.4, 16.7, 15.0, 13.9, 12.7],
diff --git a/docs/src/transforms.md b/docs/src/transforms.md
index 12a8b41..72d09a8 100644
--- a/docs/src/transforms.md
+++ b/docs/src/transforms.md
@@ -2,12 +2,14 @@
 
 A `Transform` defines a transformation of data for feature engineering purposes.
 Some examples are scaling, periodic functions, linear combination, and one-hot encoding.
+Transforms can be stateless, for example the power transform, or they can be stateful and fit to the data, such as the [`StandardScaling`](@ref).
 
 ```@meta
 DocTestSetup = quote
     using DataFrames
     using Dates
     using FeatureTransforms
+    using FeatureTransforms: fit!
 end
 ```
 
@@ -20,6 +22,15 @@ For example, the following defines a squaring operation (i.e. raise to the power
 julia> p = Power(2);
 ```
 
+A stateful transform, such as a [`StandardScaling`](@ref) should also be fit to the data before it is applied:
+```julia-repl
+julia> s = StandardScaling();
+
+julia> x = rand(5);
+
+julia> FeatureTransforms.fit(s, x);
+```
+
 ## Methods to apply a transform
 
 Given some data `x`, there are three main methods to apply a transform.
@@ -147,7 +158,7 @@ julia> M
 julia> normalize_row = StandardScaling();
 
 julia> fit!(normalize_row, M; dims=1, inds=[2])
-StandardScaling(3.0, 2.8284271247461903, true)
+StandardScaling(3.0, 2.8284271247461903)
 
 julia> normalize_row(M; dims=1, inds=[2])
 1×2 Matrix{Float64}:
@@ -156,7 +167,7 @@ julia> normalize_row(M; dims=1, inds=[2])
 julia> normalize_col = StandardScaling();
 
 julia> fit!(normalize_col, M; dims=2, inds=[2])
-StandardScaling(5.0, 1.0, true)
+StandardScaling(5.0, 1.0)
 
 julia> normalize_col(M; dims=2, inds=[2])
 3×1 Matrix{Float64}: