update

Author: spatialthoughts

docs/end-to-end-gee-supplement.html

@@ -4185,8 +4185,8 @@ <h2>Medoid Composites</h2>
 <p>Medoid, also known as Geometric Median or Multi Dimensional Median is
 often used to get a more representative sample from multi-dimensional
 data. For multispectral satellite images - a medoid composite can result
-in a better output than a simple median composite. The method was
-originally described in the paper <a
+in a better output than a simple median composite. The method and its
+advantages are described in the paper <a
 href="https://www.mdpi.com/2072-4292/5/12/6481">Seasonal Composite
 Landsat TM/ETM+ Images Using the Medoid</a>.</p>
 <p>Some key differences between a simple median composite vs. a medoid
@@ -4199,8 +4199,8 @@ <h2>Medoid Composites</h2>
 <li>In a median composite image, each band value may come from different
 images. Medoid values are guaranteed to come from the same image.</li>
 <li>Median composites may have values that are not present in the
-original images. i.e. Median of [1, 2] –&gt; 1.5. Medoid will always
-pick values present in the input data.</li>
+original images. i.e. Median of [1, 2, 3, 4] –&gt; 2.5. Medoid will
+always pick values present in the input data.</li>
 </ol>
 <p>For most users, the difference is minimal and a simple median
 composite will be adequate. However, median composites can sometimes

docs/images/introduction_to_qgis/earthquakes_part1_14.png

@@ -346,13 +346,13 @@ knitr::include_graphics('images/end_to_end_gee/alamatti_animation.gif')
 
 ## Medoid Composites
 
-Medoid, also known as Geometric Median or Multi Dimensional Median is often used to get a more representative sample from multi-dimensional data. For multispectral satellite images - a medoid composite can result in a better output than a simple median composite. The method was originally described in the paper [Seasonal Composite Landsat TM/ETM+ Images Using the Medoid](https://www.mdpi.com/2072-4292/5/12/6481).
+Medoid, also known as Geometric Median or Multi Dimensional Median is often used to get a more representative sample from multi-dimensional data. For multispectral satellite images - a medoid composite can result in a better output than a simple median composite. The method and its advantages are described in the paper [Seasonal Composite Landsat TM/ETM+ Images Using the Medoid](https://www.mdpi.com/2072-4292/5/12/6481).
 
 Some key differences between a simple median composite vs. a medoid composite:
 
 1. Median values are calculated independently for each band while Medoid value is computed by comparing pixel values across all bands and selecting the values that minimize the difference  between the median and the chosen value.
 2. In a median composite image, each band value may come from different images. Medoid values are guaranteed to come from the same image.
-3. Median composites may have values that are not present in the original images. i.e. Median of [1, 2] --> 1.5. Medoid will always pick values present in the input data.
+3. Median composites may have values that are not present in the original images. i.e. Median of [1, 2, 3, 4] --> 2.5. Medoid will always pick values present in the input data.
 
 For most users, the difference is minimal and a simple median composite will be adequate. However, median composites can sometimes produce outliers when computing spectral indices (like NDVI) whereas medoid composites are more robust against such outliers.