#* masked_coherence_by_dp_topic:
#*   attr:
#*     fillcolor: '3'
#*   desc: Psuedorandomly mask site IDs in site topic coherence scores to match other
#*     psuedorandom IDs.
#*   ext: py
#*   inputs:
#*   - coherence_by_dp_topic
#* 

import pyspark.sql.functions as F
from pyspark.sql.types import IntegerType,BooleanType,DateType,StringType

def masked_coherence_by_dp_topic(coherence_by_dp_topic):
    enabled = True
    if enabled:
        munged = coherence_by_dp_topic.withColumn('data_partner_id', F.sha2(F.concat(coherence_by_dp_topic.data_partner_id.cast(StringType()), F.lit('seed')), 256).substr(0, 3))
        return munged
    else:
        return coherence_by_dp_topic