StandardScaler (Pyspark) returns NaN values

Question

I am trying to build a linear regression model using Pyspark and MLib for predicting the closing price of stocks. The schema is as shown below.

root
 |-- Date