Table Explorer Schema-Driven Behavior¶

The table inspector in emzed_gui is not a generic grid only. Parts of the UI and several actions are enabled based on column groups present in the table.

The relevant implementation is in src/emzed_gui/table_explorer.

Core Rule¶

Feature detection is based on complete column groups, not on single columns in isolation.

The helper supported_postfixes() accepts both:

• base names like peakmap, mzmin, mzmax, rtmin, rtmax
• repeated groups with a shared postfix like peakmap__0, mzmin__0, mzmax__0, rtmin__0, rtmax__0

This means one row can expose multiple logical peak or chromatogram datasets.

Column Groups And Enabled Behavior¶

Peak / EIC plotting¶

Required column group:

• peakmap
• mzmin
• mzmax
• rtmin
• rtmax

Code:

• eic_col_names()
• has_peaks()
• get_plotting_data()
• setup_model_dependent_look()

Effect:

• shows the EIC plot
• shows the m/z plot
• enables peak visualization for selected rows
• exposes peakmap... entries in the postfix chooser

Spectra¶

Detected by columns whose name starts with spectra, especially spectra_ms2 with optional postfixes.

Code:

• hasSpectra()
• get_ms2_spectra()
• setup_postfix_chooser()
• spectrumChosen()

Effect:

• shows the spectra chooser
• adds row-dependent MS2 spectra entries
• lets the m/z plot display selected spectra directly

Chromatograms¶

Raw chromatogram data is represented by:

• chromatogram

with optional postfixes.

Code:

• chromatogram_col_names()
• get_chromatograms()
• setup_postfix_chooser()

Intended effect:

• chromatogram traces can be plotted
• chromatogram... entries appear in the chooser
• chromatogram-only tables should activate chromatogram-only mode

Integration result columns¶

Peak integration result columns:

• area
• rmse
• peak_shape_model
• model
• valid_model

Chromatogram integration result columns:

• area_chromatogram
• rmse_chromatogram
• peak_shape_model_chromatogram
• model_chromatogram
• valid_model_chromatogram

Code:

• integration_col_names()
• chromatogram_integration_col_names()
• setupViewForTable()
• IntegrateAction
• ChromatogramIntegrateAction

Effect:

• enables the integration widget when the table model is mutable and the required integration schema is present
• exposes integration targets like peakmap__x or chromatogram__x
• writes integration results back into the corresponding column group

Color¶

Optional column:

• color

Code:

• self.has_color = "color" in table.col_names
• get_color()

Effect:

• enables row or per-cell background coloring

Nested inspectors¶

If a cell contains nested Table or PeakMap values, the explorer treats those cells specially.

Code:

• setupViewForTable()
• handle_double_click()

Effect:

• those cells are marked non-editable
• double-click opens a nested inspector
• for PeakMap cells, the initial inspection window is derived from matching rtmin, rtmax, mzmin, mzmax columns sharing the same postfix

Other Schema-Dependent UI¶

Some controls are not tied to semantic names like peakmap, but still depend on the visible column schema:

• group selection choices are built from visible columns with a non-None display format in setup_choose_group_column_widget()
• sort field choices are built the same way in setup_sort_fields()

Current Known Issue¶

Chromatogram detection is currently inconsistent.

has_chromatograms() does not check for raw chromatogram columns. It checks only for chromatogram integration result columns such as area_chromatogram, rmse_chromatogram, model_chromatogram, and related fields.

This means a table that contains chromatogram data but not the corresponding integration result columns may fail to activate the expected chromatogram UI in setup_model_dependent_look().

There is a related issue in has_peakshape_model(): the chromatogram branch checks chromatogram integration columns twice instead of checking both raw chromatogram data and integration-result columns.

Suggested Documentation Contract¶

If this behavior is documented for users or plugin authors, the most accurate contract is:

• table explorer capabilities are enabled by recognized column groups
• those groups may appear once or multiple times with shared __postfix suffixes
• peak, spectra, chromatogram, and integration functionality are independent schema layers
• current chromatogram feature detection has a known mismatch between raw data columns and integration-result columns