Spectral line labels

SpecViz can display spectral line identifications (“Line labels”) on top of spectral data being displayed. Line identifications are supplied by line lists. There is a set of line lists already included in the distribution, and the user can read his/hers own line lists as text files formatted in Astropy’s ECSV format.

One a spectrum is displayed in the plot window, clicking on the green “Line Labels” button in the main menu bar will bring up a floating window which contains the functionality to manage the line lists.

Selecting line lists

Use the drop-down menu to select a line list. Alternatively, use the File menu button (the yellow folder) to read a line list from a user-supplied file.

Currently, the following line lists are supplied within the package and available via the drop-down menu:

Line list Number of lines Wavelength range
Atomic-Ionic 42 0.97 - 3.95 \({\mu}\)
CO 66 1.56 - 2.51 \({\mu}\)
H-He 145 0.72 - 3.74 \({\mu}\)
H2 226 1.09 - 3.99 \({\mu}\)
Common Stellar 95 1,215 - 10,938 Å
Common Nebular 51 3,430 - 7,130 Å
ILSS 25,800 2,950 - 13,160 Å
Reader-Corliss 46,646 16 - 39,985 Å
SDSS 48 1,034 - 8,660 Å

Once a line list is selected, a dialog pop-up will ask what is the wavelength range one wants to read from the list. The dialog is populated by default with the wavelength range spanned by the spectrum being displayed. Typing in new values in the dialog text fields will retain then during subsequent uses of the dialog, until they are re-typed again.

The dialog will display the actual number of lines that will be read from the list. Large numbers of lines trigger an alert, in the form of red text.

The two large lists must be handled with care, because if one attempts to read or select the entire list, some functionality may be affected adversely and become very slow. It is recommended that small wavelength ranges be used with those lists, in a way that no more than about 2,000 lines be read.

Line list management

Each line list, once read, ends up in a table that is placed under a separate tab. The tab name is the list name. Each tabbed panel contains several sections, described below:

  1. Header with descriptive information on the line list.
  2. The table itself. The exact column names and contents are list-dependent. Column headers can be clicked; that way, the column is sorted in ascending/descending order upon successive clicks. Hovering the mouse on a column heading may bring additional information on the column, such as units.
  3. Control section. This contains a number of controls to help configure the display of the selected lines in the table

Specific lines or groups of lines are selected in the table with the standard selection gestures provided by the underlying operating system. To select all lines in a table, click on the upper left corner of the table. To un-select all lines in a table, use the ‘Deselect’ button in the control section.

The ‘Color’, ‘Height’, and ‘Redshift’ controls allow the customization of the plot of the currently selected lines. ‘Height’ is interpreted as the fractional height on the plot window. ‘Redshift’ can be interpreted in either ‘z’ or ‘km/s’ units, according to the corresponding selector.

List sub-sets

When at least one line is selected in the table, clicking the ‘Create set’ button causes a new list to be built and displayed in a new tab after the ‘Original’ tab. The ‘Original’ tab always contains the entire original list. Successive use of table row selection gestures and the ‘Create set’ button, allows the creation of multiple sub sets. Sub-sets can be created from the ‘Original’ list, as well as from any other sub-set.

Each list sub-set carries its own group of display controls: color, height, and redshift. With these, one can customize the appearance of each sub-set on the plot.

The list sub-set capability can be combined with the column sorting and multiple row selection capabilities to implement hierarchical sorting.

Say, as an example, one wants to display all the Fe lines in blue color, and all the high intensity Fe lines (if intensity is provided by the line list) in red. One can sort the original table by species, select the subgroup of Fe lines, and create set #1. Next, on the #1 set, one sorts by intensity, selects all the high intensity lines, and creates set #2. One then de-selects everything in the Original set, and selects everything in sets #1 and #2. Finally, in set #1 one picks the blue color, and in set #2 the red color. Clicking ‘Draw’ will then plot everything.

Drawing the line labels

The ‘Draw’ button accomplishes the actual plotting. It works by finding all the lines that are selected in all tabs at once. Before performing the actual plotting, it will first erase all line labels left on screen by previous drawings. The ‘Erase’ button performs the same action, with no subsequent drawing.

At the left lower corner of the window, a counter keeps track of the total of lines selected at any time. The counter becomes red as a warning that a large number of lines is selected. Plotting a large number of line labels slows down the plot and zoom functionalities. The user must be aware that the response may become slow when large numbers of lines are selected.

The drawing operation includes a de-cluttering step. This achieves the dual goal of making the plot more readable, and faster to zoom/pan/rescale. The de-cluttering algorithm trades speed for cleverness, and a side effect of that is that, when plotting sets of lines at different heights on screen, some line labels may disappear even though they shouldn’t. Zooming in will eventually make all line labels visible.

Experience has shown that the perceived speed depends to a certain amount on the particular hardware and software platform the application is running on. Typically, a slow laptop can handle a couple of hundred lines with no problem. A faster, multi-core desktop can be pushed up to perhaps a thousand line labels before performance degrades significantly.

Results

The ‘Plotted’ tab will always contain the lines currently being displayed on the plot. The contents of this tab can be output to a ECSV file via the Export button on the top menu bar of the line lists window.

The file thus produced can be directly read by SpecViz via de File button (the yellow folder icon).