Avro deserialisation not working

Stream Processing
1

I have a pyspark application that consumes messages from a Kafka topic on Confluent. However during the deserialisation, it returns incorrect values

I have checked the schema, and the messages and they look fine. I am using spark version 3.3.0; this is how I deploy it

spark-submit --packages org.apache.spark:spark-sql-kafka-0-10_2.12:3.3.0,org.apache.spark:spark-avro_2.12:3.3.0 stream_ads.py

The schema

{"type":"record","name":"ctr","namespace":"com.adclicks","fields":[{"name":"ad_id","type":"string"},{"name":"timestamp","type":"string"},{"name":"clicks","type":"int"},{"name":"views","type":"int"},{"name":"costs","type":"double"}]}

Here is the snippet

with open("mock_data_schema.json", "r") as f:
    json_format_schema = f.read()

df = spark.readStream.format("kafka")\
    .options(**kafka_options)\
    .load()

df = df\
        .select(from_avro(col("value"), schema_str).alias("data"))\
       .select("data.*")

query = df.writeStream \
    .outputMode("append") \
    .format("console") \
    .option("truncate", "false") \
    .trigger(processingTime="3 second") \
    .start()\
    .awaitTermination()

This is the output

+-----+---------+------+-----+----------------------+
|ad_id|timestamp|clicks|views|costs                 |
+-----+---------+------+-----+----------------------+
|     |         |-1    |84035|1.5673984559807316E-76|
|     |         |-1    |84035|1.567405013908699E-76 |
|     |         |-1    |84035|1.5673958049080624E-76|
|     |         |-1    |84035|1.5673997634004282E-76|
+-----+---------+------+-----+----------------------+

What could be wrong?