Shared validation logic for event data across different validation contexts. Supports three validation levels with increasing strictness.
Usage
validate_event_data_core(
data,
validation_level = c("basic", "strict", "s7"),
return_errors = FALSE
)Arguments
- data
A data.frame or tibble to validate
- validation_level
Character string specifying validation level:
"basic": Basic structure and type validation
"strict": Strict validation for correlation computation
"s7": Full validation for S7 EventTable objects
- return_errors
Logical; if TRUE, returns character vector of errors instead of stopping on first error
Value
If return_errors=FALSE: invisible(TRUE) on success, stops on error. If return_errors=TRUE: NULL on success, character vector of errors on failure.
Details
Basic level:
Required columns (H1, H2, Analysis, Event) present
All columns are numeric
Hypothesis indices (H1, H2) are positive
Analysis numbers are positive
Event counts are non-negative
Strict level (includes basic plus):
H1 <= H2 for all rows (correlation computation requirement)
Unique combinations of H1, H2, Analysis
Sequential hypothesis and analysis indices starting from 1
Multiple analyses required
Diagonal entries exist for all off-diagonal entries
S7 level (includes strict plus):
Event counts non-decreasing across analyses for fixed H1, H2
Diagonal entries have Event >= corresponding off-diagonal entries
Allows H1 > H2 (more flexible than strict)