Exploring ​and ​Preparing ​your ​Data with BigQuery

Terms in this set (20)

What is one of the key reasons Google Cloud Platform can scale effectively to query large datasets?

Storage and compute power are handled separately and users don't manage their own data warehouse architecture
Storage and Compute power are separate resources on Google Cloud Platform. They are fungible resources that can scale up and down to meet querying demands.

Select the common reasons why organizations move to the cloud for big data analysis (choose all that apply):

1. Reduced cost compared to on-premise
2. Querying infrastructure is fully-managed

What are the elastic storage bins called in Google Cloud Storage?

Buckets

What is the best way to see how much data is being processed by your query in the BigQuery Web UI?

Click on the Query Validator to see how many bytes will be / are processed

What is a common way in SQL to identify duplicate records?

Use COUNT(field) in conjunction with a HAVING clause

What are the common tasks a data analyst must perform?

Ingest, Transform, Store, Analyze, Visualize

True or False: BigQuery stores the audit logs of each query ran against your datasets

True

Which of these most accurately describes the goal of BigQuery?tRUE

BigQuery is a petabyte-scale data warehouse. It is optimized for analytic workloads which equates a large amount of reading and processing data.

Which one of these three roles benefit from having a background in SQL for processing data?

Data Analyst, Data Scientist, Data Engineer,
All of the above

Answer = All of the above

What is the reason to not use SELECT * to explore your dataset?

The BigQuery preview data table feature is faster and free to preview records

Selecting all columns is an expensive operation performance-wise, especially with no filters

Selecting all columns, even with WHERE clause filters, will scan your entire dataset and incur charges for all bytes processed. This is a pitfall when returning potentially large columns (like very long string fields, e.g. Product Reviews).

All of the above (CORRECT ANSWER)

Which one of these SQL clauses cannot operate against an alias that was just defined in your SELECT statement?

WHERE

What is the core principle behind how BigQuery can effectively scale to process billions of rows of data?

Massive parallel processing of queries across distributed resources

Which of the following statements about BigQuery pricing is FALSE?

All BigQuery jobs incur charges and are billed directly in the Google Cloud Platform billing area

What is NOT one of the best practices for cost optimizing your queries?

Use LIMIT queries with WHERE clause filters to limit the amount of data scanned

Which of the following statements about Cloud Dataprep is FALSE?

Cloud Dataprep loads all of your data into the Transformer -- a view where your can quickly explore the quality of your dataset

What is a challenge of having a dataset that is tall (many rows) but thin (few columns)?

You are limited in the amount of insight you can get from a single record

True or False: Running Cloud Dataprep jobs automatically creates a new Cloud Spanner job behind-the-scenes to process your data

False

True or False: You can automatically output the results of your Cloud Dataprep job into a new BigQuery table

True

If you are analysing your dataset for completeness, what is one question you will always ask?

Have I been given the complete set of records to analyse or is this a subset?

What are the key goals of using the query validator in the BigQuery Web UI

1. View the amount of data you are going to process before running your query
2. Get insight into how to fix broken queries