From b3b8876a5a4f0d59e1fe61071e39cb59981bcbbb Mon Sep 17 00:00:00 2001
From: David Rotermund <54365609+davrot@users.noreply.github.com>
Date: Fri, 1 Dec 2023 02:22:30 +0100
Subject: [PATCH] Update README.md

Signed-off-by: David Rotermund <54365609+davrot@users.noreply.github.com>
---
 SVD_data_cleaning/README.md | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/SVD_data_cleaning/README.md b/SVD_data_cleaning/README.md
index f695c50..3457555 100644
--- a/SVD_data_cleaning/README.md
+++ b/SVD_data_cleaning/README.md
@@ -120,4 +120,6 @@ This is the original clean data and the reconstructed clean data for the first t
 
 ![figure 5](image5.png)
 
+## Disclaimer
 
+With decreasing number of channels the reconstructions of the pertubation loses quality. This is due to the fact, that the SVD can not distinguish between the common signal which was mixed in and a random common fate of the time series. In this example, where the clean data is generated from sine waves, this effect is especially strong. You should always take a close look at u[:, 0] which is the reconstructed common signal.