How normalize a list of time series features with pyspark?

My goal is to normalize a list of time series to perform a kmeans clustering

My dataset is a dataframe with hashtags as entries and column containing time serie features like:

+--------------------+--------------------+
|             hashtag|            features|
+--------------------+--------------------+
|     aldubseeyousoon|(119,[51,52,53,54...|
|     aldubsummerlove|(119,[9,10,11,12,...|
|          primaryday|[0.0,8.0,2.0,2.0,...|
|   aldubebfathersday|(119,[73,74,75,76...|
| dolceamoresacrifice|(119,[90,91,92,93...|
|       4yearswithexo|(119,[0,1,2,3,4,5...|


(Some features lists are given as sparse)

My question: If I want to normalize the dataset, is it enough to normalize for each line or do I have to do something else ?