# given (10-given.com)

Something you already accept as true or take for granted. A person might say "given the weather, we should stay inside" to mean "since we know it's raining."

## What a model may hear
- decorator for parameterized tests (Python testing with pytest): expects a list of values to run a test function multiple times
- precondition or assumption (mathematical proofs and formal logic): treats the following statement as an axiom to build reasoning upon
- input data block (Gherkin/Cucumber behavior-driven development): parses the clause as a setup step before actions and outcomes
- conditional probability notation (statistics and machine learning): reads P(A|B) as "probability of A given B" and applies Bayes' rules

## Where people and models part ways
- Says: "Given what I told you, what's next?"
  Means: Based on our conversation so far, suggest a next step
  May be taken as: searches for a Gherkin Given clause or treats prior messages as formal preconditions to be validated
  Say instead: "Based on what we discussed, what's next?"
- Says: "It's a given that prices will rise"
  Means: I assume prices will rise as obvious fact
  May be taken as: flags as statistical assumption requiring explicit prior distribution or proof obligation
  Say instead: "I assume prices will rise"
- Says: "Test it given these three cases"
  Means: Please check these three examples
  May be taken as: generates pytest @pytest.mark.parametrize or @given decorator syntax with Hypothesis strategies
  Say instead: "Please check these three specific examples"
- Says: "Given the data, predict Y"
  Means: Using this dataset, predict Y
  May be taken as: interprets as conditional probability P(Y|data) and outputs formal statistical derivation
  Say instead: "Using this dataset, build a model to predict Y"

## Tips
- Avoid "given that" when you mean "because" or "since"—models hear proof language
- Say "using these examples" instead of "given these cases" to avoid test-framework triggers
- Specify "no code" if you mean informal reasoning, not formal logic or programming
- Watch for models that echo Gherkin format back at you; restate in plain narrative if that happens
- In statistical requests, say "predict from" not "given" to stay in applied mode

## Often confused with
- given_name: a person's first name, not an assumption
- givenchy: a fashion brand, homophone only
- assuming: also a premise, but less likely to trigger test frameworks
- when: Gherkin action clause, pairs with given in BDD syntax
- then: Gherkin assertion clause, completes given-when-then triple
- hypothesis: Python property-testing library that uses given decorator