January blues?

The third Monday in January (#bluemonday) is supposedly the most depressing day of the year.  Not scientifically proven, and perhaps made up by the media – but maybe take some time out for yourself anyway and relax with a coffee and some of the best news, blogs, videos and articles we’ve read this week.

(and some jokes of course!)

Entered what I ate today into my new fitness app and it just sent an ambulance to my house.

Doctor’s office: All our records are electronic now just fill out these 12 forms.

Blogs of the Week

1.Connect Python to Oracle DBaaS database, and the VCN security

Gianni Ceresa wrote about setting up a JupyterLab sandbox in an Oracle Cloud Compute Instance not long ago. Now he wants to use it for some experiments with Machine Learning and graphs.  He asks, “How can I connect from Python (my JupyterLab notebook) to an Oracle DBaaS database? Must be easy using cx_Oracle, the Python interface for Oracle database.” …

2. Getting the most out of in-memory – part 2

Connor McDonald starts by writing, “In the previous post, I described the importance of checking V$IM_SEGMENTS to ensure that the in-memory store is populated to have confidence that you are indeed getting the benefit of the feature. But even if the store has been populated, when it comes to virtual columns, additional care is needed to make sure you can take advantage of the feature. Here is an example of how you might not be getting the in-memory benefits when you were expecting to (and how to solve it).”

3. Partial Indexing

David Kurtz begins his blog by saying, “It is not an uncommon requirement to find rows that match a rare value in a column with a small number of distinct values.  So, the distribution of values is skewed.  A typical example is a status column where an application processes newer rows that are a relatively small proportion of the table because over time the majority of rows have been processed and are at the final status.”

His conclusion is:

The advantage of this global partitioning approach is that it does not require any change to application code, and it does not involve partitioning the table. However, you will have to remember not to rebuild the unusable partitions, otherwise, they will have to be maintained as the table changes until you make them unusable again and, they will consume space that you will only get back by recreating the entire index.

NB: Partitioning is a licenced option that is only available on the Enterprise Edition of the database.

4. Oracle CEO Safra Catz: What Moving to Oracle Cloud Has Done for Us

Margaret Harrist lists in this blog some of the benefits Oracle has realised to date (as mentioned by Safra Catz):

  • Finance efficiencies
  • Better HR processes
  • Automated, data-driven sales and marketing
  • Lower IT costs

5. Machine Learning Evaluation Measures

Brendan Tierney starts with, “When developing machine learning models there is a long list of possible evaluation measures. On one hand this can be good as it gives us lots of insights into the models and be able to select the best model that meets the requirements.  In this post I’ll look at some of these evolution measures.”

The following evaluation measures are discussed:

  • R-Squared (R2)
  • Mean Squared Error (MSE)
  • Sum of Squared Error (SSE)
  • Root Mean Square (RMSE)

6. Deploying Oracle Analytics Cloud Remote Data Gateway in a Remote Private Subnet

Dayne Carley says, “This post is a step-by-step guide to installing and configuring RDG in a remote region’s private subnet on Oracle Cloud Infrastructure (OCI). The term Remote is used here to denote a different region e.g. OAC in Phoenix (PHX) and the DB and RDG in Ashburn (ASH). Although connectivity is simple over the internet, ensuring private connectivity is more complex and is the subject of this post.”

He also links to the documentation “Preparing Data in Oracle Analytics Cloud -Connect to On-premises Data Sources

7. OAC and OBI Supported data sources

This blog links to Visualizing Data and Building Reports in Oracle Analytics Cloud and Oracle Fusion Middleware Supported System Configurations

8. Important update of ORDS… release 19.4

Dimitri Gielis says, “For me, the following 3 improvements are worth the upgrade”

  • Performance of Rest APIs
  • Removal of PDF Generation Support
  • SQL Developer Web

9. With Analytics, Every Picture Tells a Story

John Hagerty opens with, “I heard the Rod Stewart 1971 classic “Every Picture Tells A Story” this week and it inspired me to jot down my thoughts on data storytelling.

The business rule of thumb has always been “Show me data in a way that’s easy to understand.” For many of us, that means producing a series of charts and graphs, sharing that data, and you’re done. Everyone will understand them, right? Isn’t a picture worth a thousand words?”

His four key takeaways are:

  1. Visuals alone don’t meet the needs of every data consumer.
  2. Make time to add narratives to your data for clearer, concise understanding.
  3. Define your story, then support it with the right data.
  4. Use augmented analytics to automate creation of narratives, then refine your approach going forward.

10. Manual upgrade to Oracle 19c (CDB/PDB)

Stefan says, “It is very cool to do everything with so called „auto tools“. If you prefer to do the Upgrade to 19c manually and step-by-step then you can follow my article and have fun, otherwise skip this blog article.

Before you start, you need to read a lot Doc-ID’s from Oracle Support.”

He then goes on to list some very important Doc-ID’s for the Upgrade.

He also invites us to check out this blog.

This week on Twitter

Oracle EPM Blogs shared Essbase 19c MaxL issue

Oracle IT Infrastructure posted:


Stories from: biconnector.comodtug.com and gotodba.com

Videos such as:

Clean backups are a MISTAKE

Oracle Machine Learning for Python, with Demos