The Library — browse and select your datasets
The Library is the table of every dataset in the project. It is where you browse, sort, flag, and select datasets, launch batch processing, and preview a dataset's overview images. A toolbar sits along the top; below it a splitter divides the dataset table (left) from a preview panel (right).
The numbered markers on the screenshot below point to each region; the legend that follows explains them. The small ? button at the far right of the toolbar (just past the selection summary) opens this page.
1
2
3
4
5
6
7
8
9
10
11
| # | Region | What it is and does |
|---|---|---|
| 1 | Process Selected | Adds the checked datasets to the parallel LocalQueue batch queue for processing. |
| 2 | Selection tools | Show selected only, Select All, Deselect All, Invert selection, and Select / Remove the highlighted rows from the checked set. |
| 3 | Flag tools | Set the flag colour (none / red / green / blue / gold) on the currently highlighted rows. |
| 4 | Group & file operations | Add and rename image groups, move highlighted datasets to another group, trash them, copy to a second project, save/load a selection list, and empty the trash. |
| 5 | Auto Selection Tool | Opens a bar that auto-checks the datasets whose chosen parameter falls in a numeric range (optionally filtered by flag colour). |
| 6 | Selection summary | A live count — "N datasets: X selected and Y highlighted". The ? help button sits just to its right. |
| 7 | Select column | The check-box for each dataset. The set of checked boxes is the selected set that Process Selected acts on. |
| 8 | Flag column | Shows each dataset's flag colour (a coloured marker), set with the flag tools (3). |
| 9 | Column headers | Every other column (Directory, specimen number, last processed, drift, raw stack, …). Click a header to sort by it (click again to reverse); drag headers to reorder them; the ⇔ toggle at the far right switches column widths between fitting the values and fitting the full names; right-click a header to show/hide individual columns. |
| 10 | Highlighted row(s) | The rows currently selected in the view (shown in blue). The flag, move and trash actions operate on these highlighted rows. |
| 11 | Preview panel | The overview images of the current dataset (dark-field, bright-field, parallax, shadow, PSF, average pattern, histograms, …), each with an Inspect button. Arrows step to the previous/next dataset, and the buttons below adjust contrast, overlay, and show the file header. |
The Library in the Modern interface style
| Part | What it does |
|---|---|
| Process button | ▶ Process N selected puts the selected (checked) datasets in the processing queue. |
| Select: All / None / Invert | Select every dataset, none, or the others (Ctrl+A, Ctrl+R, Ctrl+I). |
| Show Selected Only | While it is pressed, the table shows only the selected datasets (Ctrl+X). |
| Coloured dots | Flag the highlighted datasets: none, red, green, blue, gold (N R G B O). In the table the flag is a dot of its colour. |
| Organize, Selections | Menus of the group and file actions (new group, rename, move, copy to a second project, trash, empty trash) and of the selection (select or deselect the highlighted datasets, save and load a selection list). The shortcuts are as before. |
| Filter field | Words show the datasets whose name or values contain them all. A condition on a column — its parameter name or a part of its title, then <, >, <=, >=, = or != and a value — shows those that meet it, e.g. cf < 0.8 or defocus > -12000. Several are combined. |
| Chips | Under the toolbar, with their counts: all datasets, each flag colour, Not processed (no WASP step has run), Failed step, Changed since run (a parameter of a step has another value now than when the step ran) and Outliers. A chip with nothing to show is hidden. |
| Names | Each dataset's directory over its raw stack (pos_50.h5) in grey. |
| Steps column | The WASP steps 01–06 of each dataset as their icons: in colour done, faded not run, with an orange dot changed since it ran, with a red mark failed. Worked out from the dataset's LOGS/ and proc/ files a few datasets at a time, again whenever the Library is shown. |
| Preview | The dataset at the cursor: its name, group, raw stack and last processing, with Show Preview as Panels, Open in Processing and Show folder. The overviews of Preferences › Library Panels are chosen in a segmented control (a pop-up with ‹ › when there are more than five). An overview shows one figure large over a strip of all its figures: a click on a tile shows it large, ‹ › step through them, ☆ stars the large one for the ★ Starred overview, and its ⋯ (or a right-click) opens Inspect: maximize, viewer, Fourier Analyzer, file system. Show Preview as Panels shows the overview as its grid of panels instead, as laid out in Preferences › Library Panels: each panel has its own ⋯ button to maximize it (with the Fourier transform of a local square), and under the grid are Sigmoid contrast for MRC images, Activate Overlay and Show Header/Info. Show Preview as Large Figure returns to the large figure; the choice is remembered. Under the overview: the state of the steps and the first four values of the table. |
| Values | Column titles are the parameters' labels; numbers are rounded to about four digits (the full value in the tooltip); a value more than four median deviations from the median of its group is red with ▲; the last processing shows as a time ago (4 d ago, the date and time in the tooltip). |
Working with the table
Sorting and arranging columns
The table can be sorted by any column — click the column header to sort ascending, click again to reverse. Rows stay grouped by their image group. You can drag column headers to reorder the columns, right-click a header to show or hide individual columns, and use the ⇔ toggle at the right end of the header row to switch each column's width between "fit the values" and "fit the full parameter name". Which columns appear (and their order) is remembered per program mode and can also be configured in Preferences ▸ LibraryTable.
Highlighted vs. selected — two different things
4d distinguishes two independent notions, and the toolbar summary (6) shows both counts:
- Highlighted — the rows you have clicked in the view (normal click, or Shift/Ctrl-click for a range or multiple). Highlighting is just the current focus in the table; the flag, move and trash actions operate on the highlighted rows.
- Selected — the state of the Select check-boxes (7). This is the set that gets processed. You can build it directly by ticking boxes, or from the highlighted rows with the selection tools (2), or automatically with the Auto Selection Tool (5).
In short: highlighting is what you are looking at; selecting (checking) is what you will process.
Submitting datasets for batch processing
Tick the datasets you want in the Select column, then press Process Selected (1). The checked datasets are handed to the parallel LocalQueue queue, which runs the configured scripts on them concurrently. While the queue runs, the matching rows animate through their queued and processing states.
Showing a single image group
Double-click an image group to show only the datasets in that group. A bar above the table then names the group and the number of its datasets, and its Up button (or Ctrl+↑, ⌘+↑ on macOS) goes back up to the list of all groups, with the group you came from highlighted. Entering or leaving a group clears the highlighted rows, so the flag, move and trash actions never act on rows that are out of sight. The Select check-boxes are not affected.
Opening a dataset
Double-click a dataset to open it in the Manual Processing tab, where you can run individual scripts on it interactively and inspect its parameters, logs and result images. Single-clicking only highlights the row and updates the preview panel (11).