from pyspark.sql.window import Window
from pyspark.sql.functions import lag, round

# Create a Window specification based on the 'Date' column
windowSpec = Window.orderBy("Date")

# Calculate the lagged value of 'Total_global_active_power'
df2 = df2.withColumn("power_lag1", round(lag(col("Total_global_active_power"), 1).over(windowSpec), 2))

display(df2)