Best Practices
This section collects the conventions and operational guidance that help you build a correct, efficient, and resilient integration. Following these practices ensures the values you consume are interpreted the way they were intended, and that your client stays performant and within service limits at scale.
The guidance is organized into four areas — two focused on interpreting data correctly, and two focused on operating your integration efficiently.
In this section
Explains the difference between value_raw (as disclosed by the company) and value_normalized (Daloopa's quarterized value), and how the span field governs the relationship between them. Read this first if a value you receive doesn't match what you expected — most discrepancies are either a scale/unit difference or a cumulative-to-isolated quarterization effect, and this page shows how to tell them apart.
Explains the difference between value_raw (as disclosed by the company) and value_normalized (Daloopa's
Explains why fundamental_id is not a stable identifier for a data point, and recommends resolving a single canonical record by deduplicating on series_id + calendar_period + restated and taking the maximum fundamental_id. Also covers what a restated fundamental is, and why the same series and period can legitimately have both an original and a restated value.
Describes what to cache and for how long — company IDs, series structure, fundamentals, and documents — and how to drive refreshes from the latest_datapoint_created_at and model_updated_at signals rather than re-fetching on every run. Caching is the most effective lever for reducing call volume and keeping your integration responsive.
Documents the 120 requests per minute limit, the 429 Too Many Requests response returned when it is exceeded, and the patterns.
Defines the canonical period formats (YYYYQQ, YYYYHH, YYYYFY) and explains the critical distinction between calendar periods and fiscal periods. The year in a period string refers to the company's fiscal year, so a label like 2024Q1 may not mean January–March 2024 — this page shows how to align periods correctly, especially when comparing across companies.
At a glance
| Topic | What it helps you do |
|---|---|
| Units & Quarterization | Interpret value_raw vs value_normalized correctly using span. |
| Fundamental Uniqueness & Restatements | Why fundamental_id isn't stable, how to resolve a single canonical value, and what a restated fundamental is |
| Caching Strategy | Reduce call volume and refresh data only when it changes. |
| Rate Limits | Stay within 120 req/min and handle 429 responses gracefully. |
| Period Format Standards | Use canonical period formats and distinguish fiscal from calendar periods. |
Where to start
- New to the API? Begin with Period Format Standards and Understanding Units & Quarterization so you read values and periods correctly from day one.
- Building for production scale? Focus on Caching Strategy and Rate Limits to keep your integration efficient and within service limits.
Updated 4 days ago