Changes in version 0.1.1 (2026-04-24)                  

API Enhancements

  - 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.
  - Default column mappings in read_glint_survey_api() aligned with the
    columns the API export actually emits:
      - emp_id_col now defaults to "Employment ID" (was NULL).
      - sent_date_col now defaults to NULL (was "Survey Cycle Sent
        Date", a column the API does not include).
      - New manager_id_col parameter (default "Manager ID"). Previously
        hardcoded inside build_glint_survey() and unreachable from the
        caller. Net effect: a typical API call like
        read_glint_survey_api(mode = "daterange") now works with zero
        column-mapping arguments.
  - validate_glint_structure() no longer warns about an optional column
    when the caller explicitly passes NULL — an explicit NULL is now
    treated as "skip this concept" rather than "missing column".
  - New parse parameter on read_glint_survey_api() (default TRUE).
    Setting parse = FALSE skips the in-memory CSV parsing and returns
    the path to the saved zip. parse = FALSE requires save_zip_to to be
    set (or GLINT_SAVE_ZIP_TO); otherwise the call errors before making
    the API request.
  - 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).

                        Changes in version 0.1.0                        

Initial Release

Core Functions

  - 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

Analysis Features

  - 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

Organizational Features

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

Data Quality

  - 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

Documentation

  - Comprehensive function documentation with examples
  - 144 unit tests covering all major functionality
  - Support for glint_survey objects and raw data frames