How sassoftware/sas-education works
1. This repository is a SAS Education content library, not an application runtime. Its “architecture” is primarily a set of learning assets: notebooks, SAS programs, sample data, rendered documentation, model artifacts, dashboards, and article companions.
2. The main organizing boundary is the sas1 area, which contains the core training curriculum. It is split by language track and topic:
- SAS notebooks in .sasnb
- Python notebooks in .ipynb
- R notebooks and setup material
- topic sequences for data access, data engineering, machine learning, and model production
3. The training flow in sas1 is layered. It starts with data access and exploration, moves through combining and transforming data, then into model development, and finally model deployment. The same conceptual pipeline is repeated across SAS, Python, and R so learners can compare language implementations of the same workflow.
4. The data subsystem is explicit and file-based. Sample inputs live under sas1/data/input in multiple formats, including CSV, JSON, Parquet, and SAS datasets. Outputs are written to sas1/data/output as CSV, SAS datasets, and RData files. This makes the repository a closed-loop teaching environment where the notebooks consume local fixtures and produce reusable artifacts.
5. The modeling subsystem is also file-based. sas1/models stores saved model artifacts, including an .rda file and a serialized Python pipeline .pkl. These are the handoff points between interactive model development and later productionization examples.
6. Shared reusable logic is kept in sas1/code, with SAS helper scripts such as variable imputation and scoring functions. This acts as a small utility layer beneath the notebooks, separating reusable model logic from the instructional notebooks themselves.
7. The repository includes a documentation rendering boundary. docs/ contains HTML versions of the SAS learning content, while sas1/.../README.md and 00_View_SAS_Notebooks_as_HTML.md files indicate the source notebooks are meant to be viewed both in notebook form and as rendered web pages. In diagram terms, this is a static documentation/output branch parallel to the editable source notebooks.
8. The articles/ subtree is a separate publication-oriented subsystem. It holds companion material for specific SAS Education articles, such as:
- prime-number performance comparisons across SAS, Python, and CAS
- observability content for SAS Viya and SAS Container Runtime
- SCAPROC-related examples for SAS Viya
These folders are self-contained case studies rather than part of the main curriculum pipeline.
9. The observability content under articles/Observability is especially boundary-driven: it packages dashboard JSON for SAS Viya and SAS Container Runtime. This implies an external monitoring/telemetry dependency, but the repo stores only the dashboard definitions and supporting documentation, not the live monitoring stack itself.
10. Several external systems are referenced as teaching targets or data sources, which matters for the architecture diagram:
- SAS Viya / SAS Workbench for notebooks and execution
- CAS for distributed analytics in the prime-computation article
- Python and R runtimes with dependency lists
- Snowflake, Google Cloud Storage, and local files as example data sources
The repo is therefore a multi-runtime training repository centered on SAS, with examples that bridge to Python, R, and external data platforms.
11. sas1/zz_Backup is an archival mirror of earlier course structures and assets. It does not introduce new architecture; it preserves previous versions of the same learning content, data, code, and model artifacts. For diagramming, it is best treated as a historical backup boundary rather than an active subsystem.
12. The top-level repository also includes governance and support files: license, contribution guidance, security guidance, and CI-related metadata. These support the educational content but do not participate in the data or model flows.
13. Overall, the clearest architectural view is:
source educational notebooks/programs + sample input data -> executed training workflows -> generated outputs/models/docs -> companion article assets and dashboards.
The repo is organized around reproducible learning paths rather than around services, APIs, or a deployable application.
Main entry points: sas1 sas1 · 04_Addressing_the_Use_Case_with_the_Python_Language Python track · articles articles · docs docs
Components 19
Core curriculum
Primary SAS Education learning paths and reusable assets under sas1.
- sas1 · Main training curriculum with language tracks, data fixtures, reusable code, and models.
sas1 - SAS track · SAS-language curriculum notebooks for the main use case.
sas1/03_Addressing_the_Use_Case_with_the_SAS_Language - Python track · Python-language curriculum notebooks mirroring the SAS workflow.
sas1/04_Addressing_the_Use_Case_with_the_Python_Language - R track · R-language curriculum notebooks and runtime setup material.
sas1/05_Addressing_the_Use_Case_with_the_R_Language - Input data · Local fixture data in CSV, JSON, Parquet, and SAS dataset formats.
sas1/data/input - Output data · Notebook-generated CSV, SAS dataset, and RData artifacts used across lessons.
sas1/data/output - Shared code · Reusable SAS logic for imputation and scoring, shared beneath the notebooks.
sas1/code - Model store · Saved model artifacts used as handoff points into production examples.
sas1/models - Python setup · Python dependency installation and environment requirements for curriculum execution.
sas1/python
Rendered docs
Static HTML and companion documentation outputs.
- docs · Static HTML rendering branch for the SAS learning content.
docs
Articles
Standalone publication-oriented case studies and observability examples.
- articles · Standalone article examples and case studies separate from the main curriculum.
articles - Prime numbers · Cross-language SAS, Python, and CAS prime computation comparisons and benchmarks.
articles/Comparing-Prime-Computations-Using-the-New-CAS-Gateway-Action - Observability · Dashboard definitions and docs for SAS Viya and SAS Container Runtime observability.
articles/Observability - SCAPROC · SCAPROC-related SAS Viya examples and supporting scripts.
articles/SCAPROC-SAS-Viya
Repo support
Governance and contribution metadata.
- Repo support · Top-level contribution, security, licensing, and CI metadata.
Other components
- sas-education · Repository root for SAS training content, companion articles, and rendered outputs.
- External runtimes · Teaching targets and execution environments referenced by the content.
- External data sources · Example upstream sources referenced in the curriculum and articles.
- Monitoring stack · External observability systems whose dashboard definitions are stored in the repo.
Connections 25
- sas-education → sas1: contains · Primary curriculum boundary.
- sas-education → docs: renders · Static documentation branch.
- sas-education → articles: includes · Separate publication-oriented materials.
- sas-education → Repo support: governed by · Repository governance and support files.
- sas1 → SAS track: tracks · SAS implementation of the teaching flow.
- sas1 → Python track: tracks · Python implementation of the teaching flow.
- sas1 → R track: tracks · R implementation and setup path.
- sas1 → Input data: reads · Curriculum notebooks consume local input fixtures.
- sas1 → Output data: writes · Lessons produce reusable output artifacts.
- sas1 → Shared code: shares logic · Common helper routines for modeling workflows.
- sas1 → Model store: saves · Persisted model artifacts from training.
- Python track → Python setup: depends on · Python environment and library setup.
- SAS track → docs: mirrors · HTML renderings parallel the editable SAS notebooks.
- Python track → docs: mirrors · Rendered documentation also accompanies the notebook sources.
- R track → docs: mirrors · Rendered documentation also accompanies the notebook sources.
- Input data → External data sources: sources from · Examples reference Snowflake, GCS, and local files.
- SAS track → External runtimes: executes on · SAS Viya / Workbench teaching environment.
- Python track → External runtimes: executes on · Python runtime used in the lessons.
- R track → External runtimes: executes on · R runtime used in the lessons.
- Prime numbers → External runtimes: compares · Benchmarks across SAS, Python, and CAS.
- Observability → Monitoring stack: defines · Dashboard JSON for Viya and container runtime monitoring.
- SCAPROC → External runtimes: targets · SCAPROC examples for SAS Viya.
- articles → Prime numbers: contains · Prime computation article package.
- articles → Observability: contains · Observability article package.
- articles → SCAPROC: contains · SCAPROC article package.