NEWS

vivaglint 0.1.1 (2026-04-24)

read_glint_survey_api() now supports three export modes via a new mode parameter:
- "cycle" (the existing behavior) — single survey cycle.
- "survey" — every cycle of one survey.
- "daterange" — every survey in the experience active within the optional start/end date window.
survey_uuid, cycle_id, mode, start_date, and end_date now fall back to matching environment variables (GLINT_SURVEY_UUID, GLINT_CYCLE_ID, GLINT_MODE, GLINT_START_DATE, GLINT_END_DATE) when the argument is omitted. Explicit arguments always win over env vars.
When omitted, mode is inferred from which identifiers are populated, so existing read_glint_survey_api(survey_uuid, cycle_id) calls keep working unchanged.
Multi-CSV exports (common in "survey" and "daterange" modes) are returned as a named list of glint_survey objects keyed by CSV filename. Entries whose CSV does not fit the standard schema fall back to plain data.frame with a warning.
New save_zip_to parameter on read_glint_survey_api(). When set to a file or directory path (or via the GLINT_SAVE_ZIP_TO env var), the raw export zip Microsoft Graph returns is written to disk in addition to being parsed into R data frames. Useful for audit trails or for feeding the zip into other downstream tools. Defaults to NULL (no zip written).

read_glint_survey() - Import and validate Viva Glint CSV exports
summarize_survey() - Calculate comprehensive question-level metrics including mean, SD, response rates, and favorability percentages
get_response_dist() - Get detailed response value distributions

Multi-cycle comparisons: compare_cycles() tracks changes across survey cycles with delta calculations
Correlation analysis: get_correlations() supports Spearman, Pearson, and Kendall methods with significance testing
Factor analysis: extract_survey_factors() identifies latent constructs with multiple rotation options
Attrition prediction: analyze_attrition() links survey responses to employee turnover with risk ratios
Demographic analysis: analyze_by_attributes() aggregates results by any combination of employee attributes

aggregate_by_manager() - Roll up results by organizational hierarchy (direct reports or full tree)
pivot_long() - Reshape data for advanced analysis

Automatic validation of Glint export structure
Survey metadata extraction
Built-in favorability classifications for 2-11 point scales based on Glint standards
Minimum group size filtering for privacy protection