For your ML experiments, you need to process CSV data files. Size of your files is about 10GB each. Your training script loads the ingested data to a pandas dataframe object. During the runs, you get an “Out of memory” error. You decide to convert the files to Parquet format and process it partially, i.e. loading only the columns relevant from the modelling point of view. 
Does it solve the problem?

Answer options:

A.Yes
B.No

Answer: A
Option A is CORRECT because the data loaded from a CSV file can expand significantly when loaded into a dataframe in memory. Converting it to the columnar Parquet format is a viable solution because it enables loading selected columns which are necessary for the training process.
Option B is incorrect because using the columnar Parquet format instead of CSV can be used to optimize memory consumption, therefore it is a good solution.
Reference: 

https://docs.microsoft.com/en-us/azure/machine-learning/how-to-create-register-datasets
https://docs.microsoft.com/en-us/azure/machine-learning/concept-optimize-data-processing