In this newsletter:
● Impact of SAS support Services on your SAS environment
● Data Science - ETL Transformations for Creating a Modelling Dataset
● SAS on Containers - Part 2: An Introduction to Kubernetes● Meet Praveen Jayasankar
Impact of SAS Support Services on your SAS Environment
At OCS we advise our clients to regularly check the support.sas.com website. A lot of information is provided through the Knowledge Base & Maintenance pages on this website. Included in this information are the support levels of all SAS products. When browsing this information we came across this statement which certainly grabbed our attention - “SAS 9.4 is under a five-year grace period for Standard Support, starting on January 1, 2020”.
This support level also applies to SAS VIYA 3.5 (see SAS services policies).
In short this means that SAS v9.4 and VIYA 3.5 will no longer be under Standard support, but will receive the status of “Limited Support” as of January 1, 2025. For some of our customers Standard support is required to achieve the necessary level of software compliance. Life Science customers for example require Standard Support to achieve the required GxP validation of software.
So what does this mean for me as a SAS customer?
- Don’t worry, limited support will continue to provide essential hotfixes for your SAS system.
- It is important to check your required level of support, is GxP or other validation requirements needed?
- SAS may decide to extend the period of standard support. Even if this is not the case, you still have three years to migrate to SAS Viya 2020.1.
If you have questions or concerns please contact OCS. With our knowledge of SAS, validated systems and our experience with migrating Viya we can assess your current situation and provide you with a plan for moving forward.
A final tip: Do check the SAS website regularly (we advise twice a month) for more ‘attention grabbing’ announcements.
Data Science - ETL Transformations for Creating a Modelling Dataset
ETL is THE most important and time-consuming aspect of predictive modelling. An estimated 80% of a data scientist’s time is spent in the ETL stage of model development. Perhaps not as exciting as applying other aspects of the process but the ETL stage is critical for quality model development. The old saying is still relevant: "Garbage In Garbage Out"
ETL – Extract, Transform and Load
Real data is messy and needs to be cleansed. The raw data we need for our models will have duplicated data, null fields, inconsistent formats, data input errors and many other general data issues. ‘Clean’ data is a good start but chances are that the data is still not ready for modelling. The data needs to be structured optimally for the desired data model.
This article highlights a number of standard transformation methods which can be applied when preparing data for a modelling dataset. There is no standard methodology and the data scientist (or ETL developer in this case) needs to know how and when to perform the necessary steps.
- Reduce Size of Data
- Define a 'Target variable
- Identifying missing numeric data
- Identifying missing character data
- Delete variables
- Value Imputation
- Additional transformations
Once your data set is properly prepared, the modelling part is likely to be more successful. We will discuss these techniques in a future article. Until then happy ETL-ing!
SAS on Containers - Part 2 'An introduction to Kubernetes'
By Juan Sanchez
This second part of the series ‘SAS on Containers’ aims to provide an introduction to Kubernetes technologies and what it means for SAS users. SAS Viya 2020.1 and up (aka SAS Viya 4) will be the future for the SAS platforms within a few years, and its infrastructure is based on the Kubernetes technologies, which are Container based.
Why do we need to start looking at Kubernetes?
First and foremost if we look at the Support Levels for SAS® 9.4, SAS® Viya® 3.5, and Earlier Releases, it is stated:
- “SAS Viya products released before January 2020 (SAS Viya 3.x) are under a five-year grace period for Standard Support, starting on January 1, 2020.”
- “SAS 9.4 is under a five-year grace period for Standard Support, starting on January 1, 2020”
For us this means that, if we want to retain a SAS system on level A of Standard Support (required for Production and regulated and validated environments), it would be necessary to schedule a migration from SAS 9.4 and SAS Viya 3.x systems to SAS Viya 4, before January 1st, 2025.
Since SAS Viya 4 architecture is based on Kubernetes technologies, we have now 3 years to make up our minds and prepare for this inevitable, yet exciting, future.
Alright, so what is Kubernetes? And how will Kubernetes work in relationship with the SAS Software?
We can certainly give an introduction, but the good news is that SAS Institute has recently presented an introductory webinar “What Do I Need to Know About Kubernetes for Administering SAS® Viya®?”. This webinar took place on the 7th of December and can be watched on-demand with this link. During this webinar, Christine Vitron, Principal Technical Training Consultant at SAS Institute, covers how Kubernetes and SAS Viya intersect, especially in the areas of administration and deployment. Who else would be better than SAS to guide us through this?
In the next part of this series, we will be diving into the main challenges and benefits of Kubernetes. In the meantime, if you have any questions or concerns please contact us. With our experience with SAS migration and validated environments we can assess your current situation or provide you with a roadmap for the future.
I am Praveen Jayasankar, currently living in the Netherlands and working as a SAS Installation Consultant at OCS Consulting for the past 1.5 years. I am 26 years old, and come from India. I am from the city called Salem located in the state of Tamil Nadu, which is in the southern part of India. I master three languages - English (Official), Tamil (regional) and Sourastra (my mother tongue). I like driving, listening to music, cooking, and am a kean wildlife traveller. My favourite dish is Mutton Biriyani (spicy food).
I have completed my Bachelor of Engineering in Computer Science in Salem. Before joining OCS Consulting, I was working as a SAS Installation Consultant at SAS Hong Kong for almost a year. I have handled clients from different regions including India, Sri Lanka, South Africa, Kuwait, UAE, Hong Kong, Macau, and the Netherlands.
Prior to working as a SAS Installation expert, I was trained in Java, SQL and I am a certified Base SAS Programmer. Due to requirements in my previous company, and my interests, I switched to work as a SAS Installation Consultant. During my experience of more than 5 years, I have performed the installation of different solutions like SAS Anti-money Laundering, SAS Enterprise Governance Risk and Compliance, SAS Data Integration Studio, SAS Customer Intelligence, SAS IFRS 9 and 17(bundled with solutions like SAS Model Implementation Platform, SAS Infrastructure Risk Management), SAS Enterprise Miner, SAS Regulatory Risk Management, SAS Office Analytics, SAS Governance Compliance Manager, SAS Visual Analytics (Distributed and Non-Distributed) and SAS Viya (3.4 and 3.5). I have installed the solutions with or without HTTPS and have configured Single Sign On for various SAS Clients. Most of my clients are Banks, Government departments, Insurance companies and Pharmaceutical organisations. While I work individually, I also like to share my knowledge and learn from my team members.
I joined OCS Consulting B.V in September 2020. Here I mostly work on SAS Installation Projects and support tickets in the Managed Services ServiceDesk. Together with my OCS Managed Services colleagues I am committed to achieving the requirements of our clients and to deliver a quality service. I look forward to meeting you on a project in the future!