Skip to content

FAQ.md

Latest commit

 

History

History
30 lines (22 loc) · 1.86 KB

FAQ.md

File metadata and controls

30 lines (22 loc) · 1.86 KB
Raw

ValidateX Frequently Asked Questions (FAQ) ❓

Q1: Does ValidateX send my data to external servers?

No. ValidateX is 100% open-source and executes completely locally or within your private cloud/VPC environment. No dataset rows or sensitive telemetry are ever transmitted outside your network.

Q2: What execution engines are supported?

ValidateX supports four execution engines:

  1. Pandas (Default, zero-copy for in-memory DataFrames)
  2. Polars (High-performance multi-threaded Rust engine)
  3. PySpark (Distributed Big Data cluster execution)
  4. Push-Down SQL (Native SQLAlchemy queries for Postgres, Snowflake, BigQuery, and DuckDB)

Q3: How does the Data Quality Score work?

ValidateX computes a weighted score between 0 and 100 based on expectation severity:

  • 🔴 Critical (×3 weight): Null checks, uniqueness, column existence, row count.
  • 🟡 Warning (×2 weight): Range bounds, regex patterns, set membership.
  • 🔵 Info (×1 weight): Mean/stdev statistical checks, string lengths.

Score = 100 × (weighted_passed / weighted_total)

Q4: Can I run ValidateX in my CI/CD pipelines?

Yes! ValidateX includes a built-in GitHub Action (kaviarasanmani/ValidateX@main) and CLI command (validatex validate) that exits with code 0 on success and 1 on failure, making it ideal for gating deployments.

Q5: How do Slack and Microsoft Teams alerts work?

You can pass --slack-webhook or --teams-webhook to the CLI or call .send_slack() / .send_teams() on any ValidationResult. Notifications format rich Block Kit (Slack) or MessageCard (Teams) alerts with score breakdowns and failure details.

Q6: Can I write custom validation rules?

Yes. Decorate any class with @register_expectation and inherit from Expectation. You can implement _validate_pandas and _validate_polars hooks to add custom business logic.