Explanation
Correct code block:
transactionsDf.withColumn("cos", round(cos(degrees(transactionsDf.value)),2))
This QUESTION NO: is especially confusing because col, "cos" are so similar. Similar-looking answer options can also appear in the exam and, just like in this question, you need to pay attention to
the
details to identify what the correct answer option is.
The first answer option to throw out is the one that starts with withColumnRenamed: The QUESTION NO: speaks specifically of adding a column. The withColumnRenamed operator only renames
an
existing column, however, so you cannot use it here.
Next, you will have to decide what should be in gap 2, the first argument of transactionsDf.withColumn(). Looking at the documentation (linked below), you can find out that the first argument of
withColumn actually needs to be a string with the name of the column to be added. So, any answer that includes col("cos") as the option for gap 2 can be disregarded.
This leaves you with two possible answers. The real difference between these two answers is where the cos and degree methods are, either in gaps 3 and 4, or vice-versa. From the QUESTION
NO: you
can find out that the new column should have "the values in column value converted to degrees and having the cosine of those converted values taken". This prescribes you a clear order of
operations: First, you convert values from column value to degrees and then you take the cosine of those values. So, the inner parenthesis (gap 4) should contain the degree method and then,
logically, gap 3 holds the cos method. This leaves you with just one possible correct answer.
More info: pyspark.sql.DataFrame.withColumn — PySpark 3.1.2 documentation
Static notebook | Dynamic notebook: See test 3, QUESTION NO: 49 (Databricks import instructions)
