Search the site’s resources

Topic or keyword

Content familyAll content familiesDossierProtocolGuidePublic sourceScenarioToolPlatformServices Search

All words must match, regardless of case or accents. Search stays in the selected language.

From two characters: use arrows to choose, Enter to open, Escape to close. You can also submit the full search.

1657. Guide

      Repeated keys: predict row counts before joining

      An identifier appearing in both tables does not guarantee a one-to-one relationship. Each key’s frequencies determine the number of pairs. The teaching example uses two-string composite keys; FR and FI are markers only, not references to real records.

1658. Guide

      Excluding rows: NOT IN and NOT EXISTS can differ

      Finding keys absent from a second table requires a policy for missing values. A NULL in a subquery result can make NOT IN unknown. These expressions concern one column; the reader also distinguishes an equality that explicitly treats NULL values as comparable.

1659. Guide

      Timestamps: check the calendar before comparing strings

      Different spellings can represent the same instant. Conversely, alphabetical order can contradict time order when offsets or fractions differ. The comparator retains the input text and calculates within an explicit profile without asking the browser to infer a time zone.

1660. Guide

      Time windows: state the bounds to prevent double counting

      The [start, end) convention includes the start and excludes the end. Two windows can meet without assigning their shared boundary instant to both. An extraction must state this choice; it does not automatically make an API’s filters exclusive.

1661. Guide

      Connectivity: distinguish people, fixed subscriptions and mobile lines

      The extract covers 2021–2023 for France, Finland, Brazil and India. It contains 36 telecommunications observations and 12 reference populations. These four countries illustrate definitions and methods without forming a ranking or a representative global sample.

1662. Guide

      Choose a digital measure for the question being asked

      Similar names do not make indicators interchangeable. The selector compares six questions with the extract’s three telecommunications series. It also identifies questions none of them answers, to avoid turning subscriptions into a measure of equipment or behaviour.

1663. Guide

      Four countries: see what population weighting changes

      For the four selected countries in 2023, the simple mean of Internet-use shares is 81.188278%. The scenario using WDI populations as weights gives 64.343169%. Both are calculated here from published values, not copied from an official ITU or World Bank aggregate.

1664. Guide

      Top results: separate precision, recall and first success

      Two rankings may retrieve the same number of relevant documents while placing them at very different positions. This exercise separates quality within the first k results, coverage of the collection and the wait for the first useful result.

1665. Guide

      NDCG: state the gain and build the right ideal ranking

      Grades 1, 2 and 3 do not yet specify how relevance is valued. Compare linear and exponential gains on the same documents; the ideal ranking must come from the full collection at the same k.

1666. Guide

      Incomplete judgments: preserve unknowns and scope

      Relevance is assessed for a particular information need. A partially judged collection does not make unjudged documents useless. This separate example distinguishes judgment coverage, observed precision and possible bounds.

1667. Guide

      Repeated quotation: locate the passage with context

      Identical quotations can occur more than once. Web Annotation distinguishes the selected text from its immediate prefix and suffix. Context can resolve ambiguity; it proves neither authorship nor factual accuracy.

1668. Guide

      Quotation positions: count code points, not JavaScript units

      Start and end offsets are reproducible only when the text and counting unit stay the same. TextPositionSelector counts Unicode code points in logical order, including for right-to-left text.

1669. Guide

      Citing a passage: separate selection from source state

      A text selector answers “which passage?”; a state answers “which representation?”. Keeping both helps an evidence record or AI-summary check avoid attributing a quotation to the wrong version.

1670. Guide

      Observed proportion: report counts and a Wilson interval

      Five positive cases out of twenty give 25%, as do 250 out of 1,000. Under a suitable binomial model, the estimates do not have equal precision. This explorer separates the observed rate, confidence level and calculated limits.

1671. Guide

      Compare systems on the same cases: retain the pairs

      Two overall rates do not reveal which cases changed. A paired table classifies each case once using decisions from A and B under the same binary criterion. The exact McNemar test uses discordant pairs.

1672. Guide

      OCR: measure errors against an explicit reference and profile

      Levenshtein distance counts the minimum operations needed to turn a reference into candidate text. It can examine a transcription or excerpt, but does not assess translation faithfulness or answer truth.

1673. Guide

      Several OCR documents: sum errors and reference lengths

      Mean document rates and a corpus rate answer different questions. This reader exposes denominators, including a reference verified to be empty.

1674. Guide

      HTTP byte ranges: inclusive endpoints and selected length

      Resuming a transfer requires precise byte positions. HTTP byte offsets start at zero and both endpoints are inclusive. The calculation separates range validity, satisfiability for a known length and the server’s actual response.

1675. Guide

      Combining fragments: coverage, gaps and representation identity

      Adding received lengths can conceal an incomplete transfer: overlap counts the same positions repeatedly. Compute the union of the intervals and compare it with the intended representation. This geometric check remains separate from byte identity and provenance.

1676. Guide

      JSON Merge Patch: delete, retain or replace a value

      A partial update must identify its format. JSON Merge Patch describes values to merge into a target; JSON Patch describes operations. Their media types and treatment of null differ. The simulator lets you inspect a merge without calling an API.

1677. Guide

      JSON Patch: follow operations on the current document

      A JSON Patch list is evaluated in order. Each path addresses the document produced by previous operations, not a permanent snapshot of the initial document. These nine original walkthroughs show calculation states and stopping conditions.

1678. Guide

      SemVer: compare precedence without inventing identity

      SemVer 2.0.0 defines syntax and precedence for software versions with a declared public API. Treating them as plain strings or decimal numbers gives incorrect results. A dataset date or tag name does not automatically adopt this contract.

1679. Guide

      SemVer: connect a version increment to a declared change

      A version number cannot reveal a change’s impact by itself. For a public API already using SemVer with a positive major version, the team must first classify the change. The calculator then illustrates the corresponding increment; it performs no compatibility audit.

1680. Public source

      BnF: separate the reading room, catalogue and digital copy

      Beneath the Oval Room’s glass roof, reading tables and shelves give documentary research a physical setting. This photograph locates BnF Richelieu in Paris; it shows neither a Gallica page nor a book’s online availability. Finding a source requires moving from the place to the record and then to the exact document.

Choose a starting point

Choose a method

Start from a task: verify, compare, document or decide.

Check a source

Examine the origin, date, scope and limits of information.

Understand a term

Find definitions before applying a protocol.

How can I refine the search?

Start with the topic

Enter one or two specific terms: citation, CSV or provenance. Every word must be present, so a long sentence can exclude useful pages.

Broaden before filtering

Search all families first. Then select a guide, protocol or source according to the kind of answer you need.

Try different wording

Try a synonym or a term in the selected language. Quotes, OR and asterisks are not search operators.