diff --git a/README.txt b/README.txt
index 1c40c0f..aec90c6 100644
--- a/README.txt
+++ b/README.txt
@@ -5,14 +5,11 @@ How to get the data:
 1. scrape TrainingPeaks using scrape_trainingpeaks.py (select a date-range and workout-type bike)
 	a) for each athlete, move the training files to a subdirectory with their name in the "raw" folder
 	b) make sure everything is within the right time-range and remove duplicates by searching for (1) (2), etc.
-	c) anonymize the data of each athlete by mapping their name to a number and put this data in the "raw_anonymous" folder. The mapping can be uploaded to "mapping.xls"
 
-2. convert TrainingPeaks fit.gz files to csv using parse_fit_to_csv.sh:
+2. convert TrainingPeaks fit.gz files to csv using the bike2csv library
 	a) extract every .fit.gz file to .fit
-	b) convert fit file to csv with parse_fit_to_csv.py
+	b) convert fit file to csv
 
-3. preprocess TODO
-
-4. login at LibreView and download athlete data manually
-
-5. merge TrainingPeaks (cycling) data with LibreView (glucose) data
+3. preprocess dexcom and trainingpeaks
+	a) first steps of dexcom and trainingpeaks can be run at the same time
+	b) once you reach timezone preprocessing, make sure dexcom and trainingpeaks are at this point in the code, so all preprocessed data from dexcom and trainingpeaks can be used to obtain the final timezones