How to filter variable date In spark?

Question

I have a dataset that contains all the dates between 2010 and 2040 under this format

1/1/2010 
1/2/2010
1/3/2010
...
...
...
12/31/2040

I am using Spark to transform data where I'm trying to apply a filter that only keeps dates that are [today - 2 years, open in the future]

I literally tried using all the date manipulation functions Spark offers including df_calendar.filter(datediff(to_date(col("date"),"m/d/yyyy"),current_date()).gt(-730))

df_calendar.select("*").withColumn("datediff",datediff(to_date(col("date"),"m/d/yyyy"),current_date())).filter(col("datediff")>(-730))

val today = df_calendar.select(date_sub(current_date(),730))
df_calendar.filter((to_date(col("date"),"m/d/yyyy") > today ))

But I always end up with the same result, the dataset return all the values starting 1/1/2021, as it goes back "2 years" but not in terms of days. Notice that I also tried using the year() function and it also returns the same result, I'm seriously confused of the result that returns each time, I really need your help for this one.