예제 #1
0
def create_keywords_showlog_table(spark, table_name):
    df = create_keywords_raw_log(spark)
    df = df.withColumnRenamed('media', 'adv_type')
    df = df.withColumnRenamed('price_model', 'adv_bill_mode_cd')
    df = df.withColumnRenamed('action_time', 'show_time')
    df.printSchema()
    write_to_table(df, table_name)
예제 #2
0
    def create_matrix_table(self, data, table_name):
        schema = StructType([
            StructField("did_list", ArrayType(StringType(), False)),
            StructField("score_matrix",
                        ArrayType(ArrayType(FloatType(), False)), False),
            StructField("c1_list", ArrayType(FloatType(), False)),
            StructField("did_bucket", IntegerType(), False)
        ])

        df = self.spark.createDataFrame(
            self.spark.sparkContext.parallelize(data), schema)
        util.write_to_table(df, table_name)
예제 #3
0
def create_effective_keywords_table(spark, table_name):
    df = create_effective_keywords(spark)
    write_to_table(df, table_name)
예제 #4
0
def create_unified_log_table(spark, table_name):
    df = create_unified_log(spark)
    write_to_table(df, table_name)
예제 #5
0
def create_log_table(spark, table_name):
    df = create_cleaned_log(spark)
    write_to_table(df, table_name)
예제 #6
0
def create_persona_table(spark, table_name):
    df = create_raw_persona(spark)
    df = df.withColumnRenamed('gender', 'gender_new_dev')
    df = df.withColumnRenamed('age', 'forecast_age_dev')
    df.printSchema()
    write_to_table(df, table_name)