In Stable 2026.06, a new node – REST API definition – was added for use in decisions in SAS Intelligent Decisioning.  A REST API definition node enables a decision to send requests to and receive responses from an external API. Using a REST API definition node, you can integrate external services such as large language models (LLMs), credit bureaus, open banking services, email, and other services directly into your decision logic.  Currently, decisions that call a REST API definition node can only be published to the destinations of SAS Microanalytic Service (MAS) or SAS Cloud Analytic Services (CAS).   In this post, I will walk you through an example of using the REST API definition to call an API to perform a currency conversion as part of a decision flow in SAS Intelligent Decisioning. ... View more

RikdR's article on Decision Economics of Real-time Signals (https://communities.sas.com/t5/SAS-Communities-Library/Decision-Economics-of-Real-time-Signals/ta-p/994900) lists decision governance as a source of value: being able to show how and why a decision was made. I'd like to ask about one specific part of that: replaying a past decision with only what was known at the time. Some background. In a retrospective study, I reconstructed all 65 Federal Reserve SOMA awards in a sample of U.S. 10-year Treasury auctions from published rules and inputs, with every absolute difference below USD 100. Matching the outcome turned out to be the easy part. The hard part is proving that each input was actually available before the decision, and a full point-in-time replay is still to be validated. A simple case shows the problem. A file describes Monday's state but is only published on Wednesday. For a decision with a Tuesday cutoff, that file must not be used, even though the rule it feeds was already in force on Monday. So each decision needs four separate timestamps: rule effective date, data reference date, data availability time and decision cutoff. In SAS Intelligent Decisioning: 1. Where would you enforce the availability constraint: upstream in the data query, inside the rule conditions, or in a code node? 2. Can an execution be bound to specific versions of the decision, its rule sets and lookup tables, together with the input data as available at the original decision cutoff, to support an auditable point-in-time replay? 3. Beyond rule-fired and path tracking, what would you keep in the execution record to make that replay auditable? I'd be grateful for any experience or pointers to documentation. Piotr Pach ... View more

Error: ERROR: The SERVER= option is required for establishing a connection through the SAS/ACCESS Interface to PC Files.   code :   proc sql; connect to pcfiles(path="&path/database/SQL_DB.accdb" dbpassword=sastest); select UserID, Income format=dollar16., State from connection to pcfiles (select top 10, UserID, Income, State from customer order by Income desc); disconnect from pcfiles; quit;   I am learning sas sql 1 essential , currently at lesson 7 , there is an activity s107a01.sas . I don't know the server ! Please help . ... View more

Subject: Unexpected filter interaction issue (Drop-Down List and Date Slider) in SAS Visual Analytics Hi everyone, I am currently working with SAS Visual Analytics on version Long-Term Support 2025.09 (Release 20251209.1765255106922). I built a report that combines two types of controls: a Drop-Down List filter and a Date Slider filter. I noticed an unexpected behavior depending on the order in which the filters are applied: Problematic Scenario: When I first select a value from the Drop-Down List filter and then adjust the date range using the Date Slider, the Date Slider action completely resets the Drop-Down List selection. This happens even though there are records that satisfy both criteria simultaneously. Expected / Working Scenario: If I apply the filters in the reverse order -- adjusting the Date Slider first and then making a selection in the Drop-Down List -- everything works as expected and the filters apply correctly together. Is this a known issue or bug in this release, or is there a specific setting in the filter interaction options that I should modify to ensure they work seamlessly regardless of the selection order? Thank you in advance for your help! @VisualAnalytics  ... View more

Question for the clinical programmers here. TLF validation is usually well defined (double programming, compare). But the CSR body text also quotes a lot of numbers: p-values, percentages, LS means, hazard ratios. Who checks that those quotes match the final outputs, especially after a late rerun? Is that part of the programming QC, done by the medical writer, or a separate review step? And roughly how long does it take for a phase 3 CSR? ... View more