Reputation: 441
SPARK 1.6, SCALA, MAVEN
i have created a dataframe from RDD and trying to filter out all records where cola= null or empty string and colb = 2 or 3.
i tried something like this.
df.filter(WHERE $"COLA isnull AND COLB =02 & 03")
But unfortunately getting error as "cannot resolve 'COLA isnull where COLB =02 & 03'
Please help
Upvotes: 0
Views: 695
Reputation: 8758
Messed up the syntax in my comment above.
$"cola".isNull && $"colb".isin("02","03")
That syntax works for me (2.1 and 2.4). If 1.6 doesn't like it, try it this way:
val foo = List("H","D")
df.filter($"COLA".isNull && $"colb".isin(foo_*))
Upvotes: 1
Reputation: 196
You can try as below:
df.filter($"COLA".isNull && $"COLB".isin("02","03"))
Refer for more details: Convert SQL Case Statement into Spark
Upvotes: 0