Working with JSON and Array data in BigQuery 2.5 Reviews

Ashu R. · Reviewed over 2 years ago

Suraj K. · Reviewed over 2 years ago

Galib H. · Reviewed over 2 years ago

Yuandika A. · Reviewed over 2 years ago

Victor B. · Reviewed over 2 years ago

There could be a little bit more time for this lab

Jacek D. · Reviewed over 2 years ago

This lab need additional 30 mins for me

Prabakaran K. · Reviewed over 2 years ago

pradeep t. · Reviewed over 2 years ago

Kulud B. · Reviewed over 2 years ago

racingのデータセットの作成は、fruit_storeと同階層に作成する旨の記述が足りない。

Watanabe J. · Reviewed over 2 years ago

Thomas D. · Reviewed over 2 years ago

MARIN V. · Reviewed over 2 years ago

Nani D. · Reviewed over 2 years ago

Jussi R. · Reviewed over 2 years ago

Arulkumaran A. · Reviewed over 2 years ago

Akarsh K. · Reviewed over 2 years ago

Zahid G. · Reviewed over 2 years ago

Eissa A. · Reviewed over 2 years ago

Vijay M. · Reviewed over 2 years ago

This was probably the most engaging lab I've done so far, but it had some significant issues. I really liked the puzzles to create new queries, and I think other labs would strongly benefit from including more things like that. Step 2 of the subsection "Let's explore a dataset with STRUCTs" in "Task 5. Introduction to STRUCTs" is worded as though for an older version of BigQuery: "If it's not present already, click ADD DATA > Pin a project." it should instead say "If it's not present already, click ADD > Star a project by name" Step 3 of that subsection also has issues for this reason. In Task 5 step 8, counting the number of structs and arrays in a large schema feels like not-very-useful busywork and soured my impression of this lab. The Ctrl-F "total matches" count is unreliable (in Firefox at least) given how the Google Cloud console displays the schema information. Having to manually expand all the nested fields is also annoying. This was the longest part of the lab. In the same Task 5, the step "As you can imagine, there is an incredible amount of website session data stored for a modern ecommerce website. The main advantage of having 32 STRUCTs in a single table is it allows you to run queries like this one without having to do any JOINs:" doesn't make it clear that the query must be run to advance the lab. It should more explicitly address this. Task 8 contains an instruction to use the UNNEST() operator and a note saying "You will need to unpack both the struct and the array within the struct as data sources after your FROM clause" but this is contrary to the bigquery docs at https://cloud.google.com/bigquery/docs/arrays#flattening_arrays, which say "Note that for correlated cross joins the UNNEST operator is optional and the CROSS JOIN can be expressed as a comma cross join." In fact, most of the tasks of the lab don't require the use of UNNEST

Nicholas C. · Reviewed over 2 years ago

Zied N. · Reviewed over 2 years ago

Xuan S. · Reviewed over 2 years ago

Karla Ilian R. · Reviewed over 2 years ago

Gaurav M. · Reviewed over 2 years ago

Marisol F. · Reviewed over 2 years ago