Google Cloud’s Post

I personally find it very interesting! Integrating Gemini with Sheets and BigQuery can revolutionize data handling by enabling the generation or enrichment of multiple rows of data simultaneously. This approach not only enhances efficiency but also opens up new possibilities for data analysis and insight generation. A practical example is detailed in a recent blog post, showcasing the power of this integration. It's an honor to be in the Top 3% and titled us at Abacus Cambridge Partners as a Premium Partners of Google, Google Cloud! #Gemini #AI #BigQuery #Data #PremierPartners #Digitilization #Abacus

This is worth a read if you’d like your #BigQuery daily cost graph to look like this! https://bit.ly/3K2aPyA by Illirik Smirnov

How it feels when I'm preparing a 5-line example to showcase how #BigQuery works:

Users of #BigQuery partitioned tables around the world, rejoice! 🎉🎉🎉 Increasing the number of partitions allowed for a partitioned table is one of the oldest and most requested feature from our customers. I looked back to find the oldest such request I could track, and found a public issue tracker, now closed by Pavan Edara dating back to 2016 (closed, because it was asking to increase the default above 2K, which we did a while back 🙂 ). It is probably odd that such a seemingly minor change is worthy of celebration. But, like everything in BigQuery, changing the _default_ means a lot of hard core fundamental infrastructure support to ensure that it works for all users, all use cases, all the time! Kudos to the team working on this for a long time: Anna Yamaguchi Jiaming Qiu Jack H Liu Khanh-Nhan Nguyen Hua Zhang Jaliya Ekanayake Our PMs, this one is also for you: Dylan M. and Jagan Athreya 😀 cc: Pavan Edara (photo clip from #BigQuery release notes)

Why connect #ga4 to #bigquery? In today's Live with Littledata we discussed how to get permanent #data storage, faster reports and more.

Wohoo, now you can partition 27 years worth of data (if partitioning by date) in BigQuery, which was previously ~11 years. Partitioning helps increase performance by pruning partitions during query time which also results in lower costs. Not sure when or what to partition by, no worries, BigQuery automatically recommends this for you: https://lnkd.in/dXfpEC_e

A Brief Data Adventure: Things were looking up, we #Excelled at we did and #Postgresed along the way. But of course we hungered for more. Seeking wisdom, we sought some advice from a sagacious Mage and an adept DataExpert.io, It did require some #BigQuery maneuvers but we knew the assignment to Data(Build)Tool (dbt Labs) step by step. The voyage proved tough and is still in progress, but when we Looker back, it is entirely worth it! P.s: Hi Fivetran to all the tools/platforms that helped in the journey. Keep on #Postgresing !

One of #BigQuery's recent updates is the Jobs Explorer (available in preview). Jobs Explorer provides an overview of queries (and other jobs) ran in the project over a selected time range. Unlike digging through the logs, or querying the jobs information schema, this provides a simple UI and an easy way to sort by different metrics of individual jobs, such as Slot Time, Duration, Bytes Processed, Owner, etc. In the screenshot below you can see how I found the most expensive jobs in the project over the past 7 days (which I was responsible for). (as a reminder, with on-demand pricing each TB processed is charged at $6.25) To access Jobs Explorer navigate to: BigQuery --> Monitoring --> Jobs Explorer

💡 #BigQuery Hint #11: First Product for Each Customer 🧩 Problem: How do you find the first product each customer bought? This is an example of the classic problem where you need to get the first value from one column (product) for each group (purchases of a customer), sorted by another column (purchase date). 🔑 Solution: Use a window function in BigQuery that partitions by customer_id (creates for each customer a window containing all purchases) and sorts this window by purchase date clause. Then only select the first row for each window. 💡 Using the window function directly in the QUALIFY simplifies the query by avoiding an additional wrapping SELECT. Note: You don't need a GROUP BY. Using a window function together with the QUALIFY filter selecting only one row per window leads to a grouping effect already. ➕ Follow me for more tech bites on #BigQuery #SQL #datamodeling

The quantified LIKE operator is now generally available in #BigQuery. This simplifies writing queries that compare a field value to multiple patterns. Three quantifiers can follow the [NOT] LIKE operator ANY, SOME and ALL. ANY and SOME are synonymous and are true if a single pattern from the set matches the provided field value. ALL requires the provided field to match all the patterns from the set. Here are a few examples: Before: page_path LIKE "%blog%" AND page_path LIKE "%luka%" After: page_path LIKE ALL ("%blog%", "%luka%") Before: page_path LIKE "%blog%" OR page_path LIKE "%article%" After: page_path LIKE ANY ("%blog%", "%article%")