DAZL Documentation | Data Analytics A-to-Z Processing Language

Contents

Quick Index Pages (1)

Steps (34)

Recipes (24)

Topic Maps (18)

Examples (18)

Tutorials (6)

Reference (7)

Understanding nollejBase

general

slug: tutorial-understanding-nollejbase

What Is nollejBase?

nollejBase is your virtual data warehouse — but unlike traditional warehouses, it doesn’t store raw data. Instead, it captures knowledge about your data: metadata, summaries, metrics, and analytical results.

It works seamlessly with DAZL: DAZL pipelines process data, and nollejBase stores the insights, making it easy to reuse, analyze, and visualize your results without moving all your raw data.

Why nollejBase Matters

Traditional data warehouses require you to:

  • Extract data from multiple sources
  • Transform and load everything into a central database
  • Maintain heavy infrastructure for storage and queries

With nollejBase:

  • Your raw data stays where it lives
  • Only insights and summaries are stored
  • Analytical workflows are lightweight and flexible

The result: all the power of a warehouse without the overhead.

Core Concepts

1. Datasets

A dataset in nollejBase represents a collection of data or a virtual table.

  • Can be external (in a database, CSV, API, etc.)
  • Or internal (generated via DAZL pipelines)

Datasets can be referenced by name in DAZL scripts for transformation or analysis.

2. Metadata

nollejBase automatically stores metadata about each dataset:

  • Column names and types
  • Dimensions vs measures
  • Relationships between tables
  • Analytical properties like frequency counts, summaries, or cubes

Metadata allows DAZL to generate insights without reprocessing raw data each time.

3. Summaries & Results

When DAZL pipelines are executed, nollejBase stores the outputs:

  • Frequency tables
  • Aggregated metrics
  • Cubes and cross-tabs
  • Charts and dashboards

These outputs are ready to use in subsequent scripts, dashboards, or reports.

4. Versioning and Reuse

nollejBase supports tracking multiple versions of datasets and analytical results:

  • Keep snapshots of analyses over time
  • Reuse prior results in new pipelines
  • Combine insights across datasets without reprocessing raw data

Example: Storing a DAZL Pipeline Output

Here’s a simple example of a DAZL pipeline writing to nollejBase:

steps:
  - loadInline:
      data:
        - {region: "North", product: "Widget", price: 10, quantity: 5}
        - {region: "South", product: "Gadget", price: 15, quantity: 2}

  - calculate:
      assign:
        total: "$price * $quantity"

  - freq:
      columns: [region]

  - store:
      dataset: "sales_summary_by_region"

Explanation:

  • loadInline loads the sample dataset
  • calculate adds a new total column
  • freq counts rows per region
  • store saves the resulting summary in nollejBase as a dataset called "sales_summary_by_region"

This dataset is now ready for dashboards, reports, or further analysis.

How DAZL and nollejBase Work Together

  1. DAZL handles the logic: loading, transforming, analyzing, and visualizing.
  2. nollejBase provides the memory: storing results, summaries, and metadata.
  3. Together, they create a pipeline for knowledge, not just data.

Next Steps