xvs: The Control Panel

The xvs control panel is a collection of tabbed folders with the following tabs. Click on any of the links for information about the corresponding sub-panel (folder).

• Ops
• View
• File
• Markers
• Browse
• Label
• Mpeg
• SM

NOTES:

1. Many of the functions described below have keyboard shortcuts---typically uppercase alphabetic characters---that are given in parentheses following the panel/widget/function name. For instance, "The Ops sub-panel (O)" means that SHIFT-O will bring up the Ops sub-panel, "Auto (A)" means that SHIFT-A will initiate auto-mode, etc.
   
   
2. xvs operations and features (modes, applicators, plot styles etc.) generally apply on a per-window basis, so that, for example, distinct windows can be in distinct display modes at any given time. This behaviour can be changed by depressing the All radio-button widget, found on the main GUI, as well as the Ops, View and File sub-panels. Those functions/widgets which will be then applied to all windows are indicated with an asterix (*) below.
   
   
3. In the following, Left, Right, Middle, and Any denote clicks of the left, right, middle, and any mouse buttons respectively.

Ops: Mode Menu

• Auto* (A)
  Auto mode causes the server to automatically cycle through the various time-steps of data stored in a window---in effect animating the data in a continuous loop. The direction of cycling (Forward/Reverse) can be selected with the Direction widget. In addition, if the Break widget is set to On, then animation will pause at both the first and last datasets of a window: Middle resumes animation. At the current time there is no way to explicitly control the speed of animation.
  
  
• Step* (S)
  In Step mode, the left and right mouse buttons and/or the left and right arrow (cursor movement) keys cause the server to step through a window's datasets as follows:
  1. Right* or Right-arrow*: Step forwards
  2. Left* or Left-arrow*: Step backwards
  3. Middle*: Exit Step mode and return to previous mode.
  In addition, the following keys also control which of the active window's datasets is currently visible:
  1. Home*: Go to first dataset.
  2. End*: Go to last dataset.
  3. Page Up*: Go forwards about 1/10th the total number of datasets.
  4. Page Down*: Go backwards about 1/10th the total number of datasets.
  5. G*: Go to an arbitrary dataset, as entered in the Input type in that will pop up. This function may also be effected by typing an integer while the cursor is positioned in the active window (typing will be echoed in the GUI's status message box), followed by a G.
  
  
• Freeze* (F)
  This is the default viewing mode. In this case, the server simply displays a single dataset. Cursor coordinates, as well as the coordinates of the data point closest to the cursor may be displayed in this mode by holding down the Left mouse button.
  
  
• X Zoom* (Z)
  Allows the user to zoom-in/out in the x-coordinate space defined by the currently displayed dataset as follows:
  1. Right*: Zoom in by a factor of about 2 to a region centred on the current mouse position. The previous view is pushed onto a stack.
  2. Left*: Zoom out by popping the view stack onto the current view.
  3. Middle*: Exit X Zoom mode and return to previous mode.
  4. > (Greater-than *): Return to global view of dataset ("zoom all the way out")
  Note that while in this mode, the window limits are automatically rescaled with each zoom operation to fit the visible data at the current timestep. Upon exit of the mode, the window limits are again automatically rescaled, but this time to fit the data at all timesteps which is within the active x-range at the time of exit.
  
  
• Y Zoom
  Allows the user to zoom-in/out in the y-coordinate space defined by the currently displayed dataset as follows:
  1. Right: Zoom in by a factor of about 2 to a region centred on the current mouse position. The previous view is pushed onto a stack.
  2. Left: Zoom out by popping the view stack onto the current view.
  3. Middle: Exit Y Zoom mode and return to previous mode.
  
  
• XY Zoom
  Allows the user to zoom-in/out in the xy-coordinate space defined by the currently displayed dataset as follows:
  1. Right: Zoom in by a factor of about 2 to a region centred on the current mouse position. The previous view is pushed onto a stack.
  2. Left: Zoom out by popping the view stack onto the current view.
  3. Middle: Exit XY Zoom mode and return to previous mode.
  
  
• Edit* (E)
  Edit mode alllows the user to edit the data in the active window.

  Implementation of this mode is currently incomplete, so users should avoid its use.

Ops: Other Widgets

• All: Enables/disables the apply-to-all-windows feature.
• Select (compress)*: Using index vector notation, selects specified datasets from the active window and deletes the rest. For example, 1-*/2 will select every second dataset (starting from the first) of the active window, effectively halving the total amount of data that is stored.
• Trim* (T): Deletes the last dataset from the active window. Particularly useful if the data originated from an .sdf file whose last dataset is incomplete/corrupted due to interrupted/abnormal program execution.
• T Rev*: Reverses the storage and display order of the active window's datasets.
• D Mean*: Replaces the data in the active window with its deviation about its mean, on a per timestep basis. Particularly useful for datasets with small intrinsic variation compared to the size of the data.
• Split L: Prompts the user to enter cut x-coordinate x_c via Any click. After mouse click, creates new window containing duplicate of all data in the active window with x < x_c
• Split R: Prompts the user to enter cut x-coordinate x_c via Any click. After mouse click creates new window containing duplicate of all data in the active window with x > x_c
• Thin*: "Thins" data in active window so that line segments joining final data points deviate from those joining the initial data points by no more than approximately the angular displacement defined by the thin angle.
• Thin angle: Controls thinning operation defined above; smaller thinning angles produces less thinning.
• Reset: Resets thinning angle to startup value.
• Name (R*): Window name: the active window may re re-named by typing in this widget.
• Index*: Current display index. The dataset currently being viewed may be changed by typing in this widget.

• X min*: Sets the x_min value of the current view.
• X max*: Sets the x_max value of the current view.
• Y min*: Sets the y_min value of the current view. Implicitly enables fixed domain mode.
• Y max*: Sets the y_max value of the current view. Implicitly enables fixed domain mode.
• Domain*: Sets the domain mode to either fixed or floating. When in floating mode, the view is autoscaled to accomodate the extremal x and y values of all incoming data. In fixed mode, incoming data lying outside of the current view is clipped (discarded).
• Full scale* (W): Enables/disables full scale mode. When enabled, the view is adjusted at each timestep to match the extremal y values of the corresponding dataset. Particularly useful when animating/stepping data with a very large dynamic range. Note that the short-cut key, W, is a toggle.
• Border*: Auto x bordering enlarges the x range of the view by a fixed fractional amount. Auto xy bordering does the same for both the x and y ranges of the view.
• All: Enables/disables the apply-to-all-windows feature.
• View stack A display widget that shows the current contents of the view stack as well as the current view. The current view is highlighted in red, entries in the view stack appear after the current view, in top-down order.
• Push: Push the current view onto the view stack.
• Pop: Pop the top entry of the view stack.
• Set*: Set the current view to the top entry of the view stack.
• S & R (Set & Rescale)*: Set the current view to the top entry of the view stack, then readjusts (y_min,y_max) to coincide with the extremal y values taken over all data (all timesteps) within the x-range of the view.
• Rescale*: Recomputes the global bounding box of the data in the active window and sets the current view to the result.
• Clip* (P): Deletes all data in the active window that lies outside of the current view.
• Name (R*): Window name: the active window may re re-named by typing in this widget.
• Index*: Current display index. The dataset currently being viewed may be changed by typing in this widget.

Note that .sdf file loading can also be accomplished via the sdftoxvs command, and that data from the server can be saved as files using the xvs save ... or xvs saveall commands.

• Open: If no entries in the File Viewer have been selected (see below), this button will pop up a file browser with a filter that selects files with an .sdf extension. The file that is chosen via the browser is then loaded into xvs. Otherwise, the button operates as described in the File Viewer section.
• Save*: Saves the data in the active window to an .sdf file with the name given in the text widget to the right of the Save button, and in the directory specified in the WD text-type-in widget. The user will be prompted to confirm overwrite if the save will overwrite an existing file.
• All: Enables/disables the apply-to-all-windows feature.
• WD*: Displays the working directory associated with the active window. The working directory can be changed by typing a valid directory name in the widget.
• Csd*: If the file $HOME/sd exists and contains the name of a valid directory, set the active window's working directory to that name.
• Name (R*): Window name: the active window may re re-named by typing in this widget.
• File Viewer The display area at the bottom of the panel is a browser that displays all .sdf files in the working directory that contain rank-1 data sets. Any file that will be clobbered by a Save operation is highlighted in red. Entries in the browser area can be selected/deselected via Any mouse clicks---hold a mouse button down and sweep to select/deselect multiple entries. Once selected, the corresponding files can be loaded into xvs by clicking the Open button.

• Marker color: This pulldown menu is used to set the marker (points or polygons) color.
• Line color: This pulldown menu is used to set the line color.
• Polygon size: Sets the size of polygon markers to an integer multiple of the base size.
• Point size: Sets the size of polygon markers to an integer multiple of the base size.
• Line width: Sets the line width to an integer multiple of the base
• N polygon: Sets the number of sides for polygon markers.
• Apply to all: Apply current settings of all widgets in the panel to all windows.
• Multi mode: This pulldown menu enables and disables multi mode. When this mode is enabled, a change in sense of dx = x_j+1 - x_j triggers a change in marker and/or line color. This is particularly useful when multiple datasets with commensurate times have been sent to xvs, such as will often occur in parameter space surveys. In such cases, xvs simply appends the new data at a given time to the old data at that same time, and then with multi mode enabled, these logically distinct datasets will be displayed with distinct marker and/or line colors.
• Tag by dx: Add a "color offset" to the marker and/or line color which is a function of the local scale of discretization, dx. Used to visualize results from locally uniform adaptive mesh refinement calculations, whereby data values will be color-coded according to dx.
• Tag by const.: Add a constant "color offset" to the marker and/or line color. This function is rarely used.
• Tag delete: Delete any tag data associated with the current window's data.
• Auto/dx on: Enable feature whereby incoming data is automatically tagged by local scale of discretization.
• Auto/dx off: Disable auto/dx tagging.

1. Name (R*): Window name: the active window may re re-named by typing in this widget.
2. Index: Current display index. The dataset currently being viewed may be changed by typing in this widget.
3. Format: This type-in widget defines the printf-style format specification used in the browsing output window. The format may be changed by typing a valid format code for a double in the widget.
4. Reset: This button resets the display format to the xvsstart-up value.

• Label: This radio button widget is used to select which annotation is to be edited with the other widgets. Available annotations are as follows:
  • Name: Window name.
  • Time: Dataset time.
  • Index: Dataset index and maximal index (number of datasets in window).
  • Limits UC: Upper right corner limits: (x_min,y_min)
  • Limits YC: Lower left corner limits: (x_max,y_max)
  • Display mode: Current mode (Freeze, Auto, Step, etc.); shown in lower right corner.
  • Mouse position: Coordinates of mouse and coordinates of data point nearest to the mouse mouse (displayed when the Middle mouse button is held down in the active window.)
• Display: This radio button controls the visibility (on/off) of the selected annotation.
• Format: This text type-in defines the printf-style format code used to format the selected annotation.
• Color: This pull-down menu sets the color of the selected annotation.
• All on: This button makes all annotations visible.
• All off: This button makes all annotations invisible.
• Reset: This button resets widget settings for the selected annotation to their start-up values.
• Reset all: This button resets widget settings of all annotations to their startup values.
• Lock limits: When enabled, changes made to either one of Limits UC or Limits LC are automatically applied to the other.

The Berkeley MPEG tools, including executables mpeg_encode and mpeg_play are required for MPEG generation and playback. These tools may be downloaded via THIS PAGE.

• Name (R*): Window name: the active window may re re-named by typing in this widget.
• Record: Pushing this button will cause the generation of an MPEG movie from the active window---the movie will be saved as a file with an .mpeg extension in the directory defined by the WD type-in. The GUI may resize slightly prior to the recording operation due to restriction on the pixel widths of images which are to be MPEG-encoded. The user will be prompted to confirm overwrite if the generated MPEG file will clobber an existing file. Depending on the number of frames (timesteps) being recorded, and the complexity of each frame, the generation of the MPEG movie may take a significant amount of time (minutes), and xvs will echo progress messages to standard output as the recording proceeds.
• Playback: Playback the most recently generated MPEG movie using the command displayed in the Playback command text widget.
• WD: This text type-in defines the working directory for MPEG generation and storage. The WD may be set by typing in a valid directory name in the widget.
• Size: Displays the width and height, in pixels, of the images that will make up the MPEG movie (xvs display windows always have a unit aspect ratio).

• WD: This text type-in defines the working directory for sm interaction. The WD may be set by typing in a valid directory name in the widget.
• Gen.: Pressing this button causes the data defined at the current display time in the active window to be dumped to a file with a .dat extension in the working directory. A template sm script is then generated and saved in a file named sm_NAME---where NAME is replaced with the active window name (only alphanumeric characters and underscores survive), also in that directory. xvs then spawns two shell windows, one runs an sm session, into which sm_NAME has been read, and the other invokes the text editor defined by the EDITOR environment variable on sm_NAME. If the EDITOR environment variable is not defined, vi is used.
• Gen. PS (lpr): As above, but have sm generate a Postscript verision of the plot and send it to the system default lineprinter.
• Csd: If the file $HOME/sd exists and contains the name of a valid directory, set the Sm working directory to that name.
• Cd : Set the Sm working directory to the xvs default: /tmp/sm_$USER
• Refresh: Update the display area at the bottom of the panel, which lists all files in the sm working directory.
• X name: This text-type in defines the string which will be used to label the x-axis in the template sm script file.
• Y name: This text-type in defines the string which will be used to label the y-axis in the template sm script file.

CAUTION!: Ensure that all of the shells (child processes) which are created via Gen. are closed (by, e.g., exiting the sm and editor sessions), prior to exiting xvs. Otherwise, network resources associated with xvs, including the main communications port, will remain open, and xvs will not subsequently re-start until the child processes have been terminated.