Abstract -------- GLE is a high quality graphics package for scientists, combining a user friendly interface with a full range of facilities for producing publication quality graphs, diagrams, posters and slides. GLE provides LaTeXquality fonts together with a flexible graphics module which allows the user to specify any feature of a graph (down to the line width of the subticks, for example) Complex pictures can be drawn with user defined subroutines and simple looping structures. Current device drivers support DECWINDOWS, REGIS, TEK4010, all PC graphics cards, VT100s, HP-Plotters, PostScript Printers, EPSON Printers and LaserJet Printers. GLE runs on both VAXes and PCs, giving an identical user interface on both platforms. GLE Installation on a PC ------------------------ To install GLE, put the distribution disk into drive a: and type: a:install GLE requires at least 800K of disk space for a minimal installation and 3.3M for a full installation (including all device drivers and fonts). GLE also requires 530K or more of free memory. Use the DOS command CHKDSK to check this figure. If you don't have enough then take copies of your AUTOEXEC.BAT and CONFIG.SYS files and then remove as much as possible from these files. GLE may work with less memory depending on what you are drawing. The installation disk contains a version of CGLE which will make use of epanded memory. This version can run with 70K less memory but if you don't have expanded memory then it has to use your hard disk instead which is a great deal slower. Running GLE ----------- To get GLE running interactively you must be on a VAXstation, graphics terminal, or PC. If you are using a PC you should make certain you have an up to date copy of GLE (ask via VAX-mail to GRV::SRGHCXP or InterNET srghcxp@grv.dsir.govt.nz) Running GLE on a PC ------------------- The command to run GLE is: C:\GLE> cgle myfile.gle Note: the command is `CGLE' not `GLE' but the directory is `GLE', this is for historical reasons. To print a GLE file to the laser printer you would type: C:> psgle myfile C:> print myfile.ps To produce an .eps file for inclusion in WordPerfect you would type: C:> psgle myfile /eps This will create a file called myfile.EPS. If your PC is connected to a VAX computer which has a PostScript printer you may copy MYFILE.PS to the VAX using a standard file transfer program (e.g. FTP, KERMIT, VDISK) The best way to see what GLE can do is to have a play with it, simply start it up and try out some of the examples: Press F3 (Load file) Press (for a menu of GLE files) Use arrow keys to select example, then press Press F10 to draw the picture Press ESC to get back to the GLE editor. ... When you find a graph try pressing F9 and modifying one of the fields (use F1 for an explanation of each field). Running GLE on a VAX -------------------- The command to run GLE is: $ cgle myfile.gle $ cgle myfile.gle /dev=regis $ cgle myfile.gle /dev=x See the directory CGLE_EXAMPLES: for examples and templates. To get access to these files from the GLE menus use the commands: $ define workarea sys$login:,cgle_examples: $ set default workarea: $ cgle stack4b.gle ! or any other example If your keyboard doesn't have the function keys F9 thru F14 you can use GOLD (PF1) followed by the numbers 9,0,1,2,3,4 (on the top of the QWERTYUIOP keypad). Keyboard Mappings: VT100 VT200 PC Meaning GOLD 1 F11 F1 Help GOLD 2 F12 F2 Save GOLD 3 F13 F3 Load GOLD 4 F14 F4 Save-as GOLD 9 F9 F9 Graph-menu GOLD 0 F10 F10 Draw-it Control+Z Control+Z Control+Z Exit/Escape Alt+X Exit/Escape Control+E Control+E Calls VAX EDT Control+F Control+F Toggle fast/slow text Control+R Control+R F5 Shows errors Control+S Shells to DOS Supported devices: VT100, REGIS (VT125, VT240), TEK4010, VWS, XWindows. Supported output: PostScript, HPGL, Epson, Epson 24pin, HP Deskjet. To create a PostScript output file (.PS) and automatically print it to the LASER queue you would type: $ cgle myfile /print To produce an .eps file for inclusion in LaTeXyou would type: $ cgle myfile /dev=eps To produce a .ps file suitable for printing to a laser writer type: $ cgle myfile /dev=ps It didn't work, bugs!!! ----------------------- If the installation fails, or one of the example GLE files fails to work then the most likely reason is a shortage of memory due to too many memory resident programs/drivers. To fix this remove these utilities from your autoexec.bat and config.sys files temporarily. There may well be a bug in your GLE file, try using the trace option to find the bug. On a PC: C:\GLE> psgle myfile /trace On a VAX: $ cgle myfile /dev=ps /trace Another reason for a failure is a bug in GLE, Please report bugs to Chris Pugmire (in%"srghcxp@grv.dsir.govt.nz", grv::srghcxp) so they can be fixed. If possible, try and find a way of repeating the problem, then send relevant GLE and data files with an outline of what is wrong. Drawing a Line on a Page ------------------------ Let's start with drawing a line on the page. GLE has to know what size piece of paper you are working with. You can tell it by giving a size command: size 18 27 This specifies a piece of paper 18cm wide and 27cm high. Now you must define a current point by moving to somewhere on the page: amove 2 4 The origin (0,0) is at the bottom left hand corner of the page. Now suppose we wish to draw a line from this point across 1 cm and up 2 cm: size 18 27 amove 2 4 rline 1 2 That was a relative movement as the x and y values were given as distances from the current point, alternatively we could have used absolute coordinates: size 18 27 amove 2 4 aline 3 6 ! absolute. Now if you want to draw some text on this page at the current point you would use the text command: text Hi there So we have now constructed an entire GLE program as follows, with the results illustrated below: size 12 8 box amove 2 4 rline 1 2 text Hi there Drawing a Simple Graph ---------------------- This section will describe how to go about drawing a simple graph. The following data points are contained in a file called TUT.DAT x y ----- (These top two lines do not appear in the file) 1 2 2 6 3 2 4 5 5 9 The data is in two columns with white space separating each column of numbers. The following commands will draw a simple line graph of the data. size 6 3 begin graph size 6 3 data tut.dat yaxis min 0 d1 line marker star msize .2 end graph The first size command defines the size of the piece of paper which may contain several graphs. The second defines the size of this particular graph. The actual graph axes are by default 0.7 of these dimensions. The ratio can be changed with the vscale and hscale commands. Without the "yaxis min 0" command the yaxis would start at 2.0, because that is the minimum y value in the data. This makes the graph easier to interpret. Changing the above d1 command to: d1 line marker circle will mark each data point with a circle instead of a star. See appendix A for a list of other marker names. A smooth line can be drawn between the data points by changing the d1 command to: d1 line marker circle smooth The order of the commands is not important, except that circle is a parameter for the qualifier marker and therefore must come straight after it. E.g., d1 [marker circle] [line smooth] [color blue] GLE Primitives -------------- A GLE command is a sequence of keywords and values separated by white space (one or more spaces or tabs). Each command must begin on a new line. Keywords may not be abbreviated, the case is not significant. All coordinates are expressed in centimetres from the bottom left corner of the page. GLE uses the concept of a current point which most commands use. For example, the command aline 2 3 will draw a line from the current point to the coordinates (2,3). The current graphics state also includes other settings like line width, color, font, 2d transformation matrix. All of these can be set with various GLE commands. Graphics Primitives (a summary) ------------------------------- ! comment @ xxx aline x y [arrow start] [arrow end] [arrow both] amove x y arc radius a1 a2 arcto x1 y1 x2 y2 rad begin box [fill pattern] [add gap] [nobox] [name xyz] begin clip begin origin begin path [stroke] [fill pattern] [clip] begin rotate angle begin scale x y begin table begin text [width exp] begin translate x y bezier x1 y1 x2 y2 x3 y3 bigfile filename.gle box x y [justify jtype] [fill color] [name xxx] [nobox] circle radius [fill pattern] closepath curve ix iy [ x1 y1 x y x y ... xn yn] ex ey define marker markername subroutine-name for var = exp1 to exp2 [step exp3] command [...] next var grestore gsave if exp then command [...] else command [...] end if include filename join object1.just sep object2.just marker marker-name [ scale-factor] postscript filename.eps width-exp height-exp rbezier x1 y1 x2 y2 x3 y3 return exp reverse rline x y [arrow end] [arrow start] [arrow both] rmove x y save objectname set cap butt | round | square set color col set dashlen dashlen-exp set font font-name set fontlwidth line-width set hei character-size set join mitre | round | bevel set just left | center | right | tl | etc... set lstyle line-style set lwidth line-width sub sub-name paramter1 paramter2 etc text unquoted-text-string write string$ Graphics Primitives (in detail) ------------------------------- ! comment Indicates the start of a comment. GLE ignores everything from the exclamation point to the end of the line. @ xxx Executes subroutine xxx. aline x y [arrow start] [arrow end] [arrow both] Draws a line from the current point to the absolute coordinates (x,y), which then becomes the new current point. The arrow qualifiers are optional, they draw arrows at the start or end of the line, the size of the arrow is proportional to the current font height. amove x y Changes the current point to the absolute coordinates (x,y). arc radius a1 a2 Draws an arc of a circle in the anti-clockwise direction, centered at the current point, of radius radius, starting at angle a1 and finishing at angle a2. Angles are specified in degrees. Zero degrees is at three o'clock and Ninety degrees is at twelve o'clock. arc 1.2 20 45 The command narc is identical but draws the arc in the clockwise direction. This is important when constructing a path. arcto x1 y1 x2 y2 rad Draws a line from the current point to (x1,y1) then to (x2,y2) but fits an arc of radius rad joining the two vectors instead of a vertex at the point (x1,y1). begin block_name ... end block_name There are several block structured commands in GLE. Each begin must have a matching end. Blocks which change the current graphics state (e.g. scale, rotate, clip etc) will restore whatever they change at the end of the block. Indentation is optional but should be used to make the GLE program easier to read. begin box [fill pattern] [add gap] [nobox] [name xyz] Draws a box around everything between begin box and end box. The option add adds a margin of margin cm to each side of the box to make the box slightly larger than the area defined by the graphics primitives in the begin box end box group (to leave a gap around text for example). The option nobox stops the box outline from being drawn. The name option saves the coordinates of the box for later use with the join command. begin clip This saves the current clipping region. A clipping region is an arbitrary path made from lines and curves which defines the area on which drawing can occur. This is used to undo the effect of a clipping region defined with the begin path command. See the example CLIP.GLE in appendix B at the end of the manual. begin origin This makes the current point the origin. This is good for subroutines or something which has been drawn using amove,aline. Everything between the begin origin and end origin can be moved as one unit. The current point is also saved and restored. begin path [stroke] [fill pattern] [clip] Initialises the drawing of a filled shape. All the lines and curves generated until the next end path command will be stored and then used to draw the shape. stroke draws the outline of the shape, fill paints the inside of the shape in the given color and clip defines the shape as a clipping region for all future drawing. Clipping and filling will only work on PostScript devices. begin rotate angle The coordinate system is rotated anti-clockwise about the current point by the angle angle (in degrees). For example, to draw a line of text running vertically up the page (as a Y axis label, say), type: begin scale x y Everything between the begin and end is scaled by the factors x and y. E.g., scale 2 3 would make the picture twice as wide and three times higher. begin table This module is an alternative to the TEXT module. It reads the spaces and tabs in the source file and aligns the words accordingly. A single space between two words is treated as a real space, not an alignment space. With a proportionally spaced font columns will line up on the left hand side but not on the right hand side. However with a fixed pitch font, like tt, everything will line up. begin text [width exp] This module displays multiple lines/paragraphs of text. The block of text is justified according to the current justify setting. See the set just command for a description of justification settings. If a width is specified the text is wrapped and justified to the given width. If a width is not given, each line of text is drawn as it appears in the file. Remember that GLE treats text in the same way that LaTeXdoes, so multiple spaces are ignored and some characters have special meaning. E.g, \ ^ _ & { } To include Greek characters in the middle of text use a backslash followed by the name of the character. E.g., 3.3\Omega S would produce ``3.3OS''. To put a space between the Omega and the S add a backslash space at the end. E.g., 3.3\Omega\ S produces ``3.3O S'' Sometimes the space control characters (e.g. \:) are also ignored, this may happen at the beginning of a line of text. In this case use the control sequence \glasswhich will trick GLE into thinking it isn't at the beginning of a line. E.g., text \glass \:\: Indented text There are several LaTeXlike commands which can be used within text, they are: ^{} Superscript _{} Subscript \\ Forced Newline \_ Underscore character \, .5em (em = width of the letter `m') \: 1em space \; 2em space \char{22} Any character in current font \chardef{a}{hello} Define a character as a macro \def\v{hello} Defines a macro \movexy{2}{3} Moves the current text point \glass Makes move/space work on beginning of line \rule{2}{4} Draws a filled in box, 2cm by 4cm \setfont{rmb} Sets the current text font \sethei{.3} Sets the font height (in cm) \setstretch{2} Scales the quantity of glue between words \lineskip{.1} Sets the default distance between lines of text \linegap{-1} Sets the minimum required gap between lines begin translate x y Everything between the begin and end is moved x units to the right and y units up. bezier x1 y1 x2 y2 x3 y3 Draws a Bezier cubic section from the current point to the point (x3,y3) with Bezier cubic control points at the coordinates (x1,y1) and (x2,y2). For a full explanation of Bezier curves see the PostScript Language Reference Manual. bigfile filename.gle This command reads the file one line at a time, compiles each line and executes it. This means it can read any sized file. However, complex multi-line commands cannot be used. Subroutines can be used but not defined, inside the bigfile. Note: there is also a bigfile option in the graphing module for large datasets. box x y [justify jtype] [fill color] [name xxx] [nobox] Draws a box, of width x and height y, with its bottom left corner at the current point. If the justify option is used, the box will be positioned relative to the specified point. E.g., TL = top left, CC = center center, BL = bottom left, CENTRE = bottom centre, RIGHT = bottom right, LEFT = bottom left. See set just for a description of justification settings. If a fill pattern is specified, the box will be filled. Remember that white fill is different from no fill pattern - white fill will erase anything that was inside the box. circle radius [fill pattern] Draws a circle at the current point, with radius radius. If a fill pattern is specified the circle will be filled. closepath Joins the beginning of a line to the end of a line. I.e., it does an aline to the end of the last amove. curve ix iy [ x1 y1 x y x y ... xn yn] ex ey Draws a curve starting at the current point and passing through the points (x1,y1) (xn,yn), with an initial slope of (ix,iy) to (x1,y1) and a final slope of (ex,ey). All the vectors are relative movements from the vector before. define marker markername subroutine-name This defines a new marker called markername which will call the subroutine subroutine-name whenever it is used. It passes two parameters, the first is the requested size of the marker and the second is a value from a secondary dataset which can be used to vary size or rotation of a marker for each point plotted. for var = exp1 to exp2 [step exp3] command [...] next var The for ... next structure lets you repeat a block of statements a number of times. GLE sets var equal to exp1 and then repeats the following steps. If var is greater than exp2 then GLE commands are skipped until the line after the next statement. The value exp3 is added to var. The statements between the for and next statement are executed. itemize If exp1 is greater than exp2 then the loop is not executed. grestore Restores the most recently saved graphics state. This is the simplest way to restore complicated transformations such as rotations and translations. It must be paired with a previous gsave command. gsave Saves the current graphics transformation matrix and the current point and the current color, font etc. if expression then command [...] else command [...] end if If expression evaluates to true, then execution continues with the statements up to the corresponding else, otherwise the statements following the else and up to the corresponding end if are executed. amove 3 3 if xpos()=3 then text We are at x=3 else text We are elsewhere end if Note: end if is not spelt endif. include filename The commands in filename are read just as if they were in the current GLE file. With a large include file GLE may run out of memory. If this happens, use the bigfile command instead of include. Note: there is also a bigfile option in the graphing module. join object1.just sep object2.just Draws a line between two named objects. An object is simply a point or a box which was given a name when it was drawn. The justify qualifiers are the standard GLE justification abbreviations (e.g. TL=top left, see set just for details) If sep is written as -, a line is drawn between the named objects e.g. join fred.tr - mary.tl Arrow heads can be included at both ends of the line by writing sep as <->. Single arrow heads are produced by <- and ->. Note that sep must be separated from object1.just and object2.just by white space. If the justification qualifiers are omitted, a line will be drawn between the centres of the two objects (clipped at the edges of the rectangles which define the objects). See Chapter 5 for an example of joining objects. marker marker-name [ scale-factor] Draws marker marker-name at the current point. The size of the marker is proportional to the current font size, scaled by the value of scale-factor if present. Markers are referred to by name, eg. square, diamond, triangle and fcircle. Markers beginning with the letter f are usually filled variants. Markers beginning with w are filled with white so lines are not visible through the marker. For a complete list of markers refer to Appendix A.1. postscript filename.eps width-exp height-exp Includes an encapsulated postscript file into a GLE picture, the postscript picture will be scaled up or down to fit the width given. On the screen you will just see a rectangle. Only the width-exp is used to scale the picture so that the aspect ratio is maintained. The height is only used to display a rectangle of the right size on the screen. rbezier x1 y1 x2 y2 x3 y3 This command is identical to the BEZIER command except that the points are all relative to the current point. return exp The return command is used inside subroutines to return a value. reverse Reverses the direction of the current path. This is used when filling multiple paths in order that the Non-Zero Winding Rule will know which part of the path is `inside'. With the Non-Zero Winding Rule an imaginary line is drawn through the object. Every time a line of the object crosses it from left to right, one is added to the counter; every time a line of the object crosses it from right to left, one is subtracted from the counter. Everywhere the counter is non-zero is considered to be the `inside' of the drawing and is filled. rline x y [arrow end] [arrow start] [arrow both] Draws a line from the current point to the relative coordinates (x,y), which then become the new current point. If the current point is (5,5) then rline 3 -2 is equivalent to aline 8 3. The optional qualifiers on the end of the command will draw arrows at one or both ends of the line, the size of the arrow head is proportional to the current font size. rmove x y Changes the current point to the relative coordinate (x,y). If the current point is (5,5) then rmove 3 -2 is equivalent to amove 8 3. save objectname This command saves a point for later use with the join command. set cap butt | round | square Defines what happens at the end of a wide line. set color col Sets the current color for all future drawing operations. There are several pre-defined colors which can be specified by name. black, white, red, green, blue, cyan, magenta, yellow, grey10, grey20 ... grey90, shade1 ... shade5, grid1 ... grid5 It is also possible to specify a grey scale as an expression with 0.0 = black and 1.0 = white. set dashlen dashlen-exp Sets the length of the smallest dash used for the line styles. This command MUST come before the set lstyle command. This may be needed when scaling a drawing by a large factor. set font font-name Sets the current font to font-name. Valid font-names are listed in Appendix A.2. There are three types of font: PostScript, LaTeXand Plotter. They will all work on any device, however LaTeXfonts are drawn in outline on a plotter, and so may not look very nice. PostScript fonts will be emulated by LaTeXfonts on non-PostScript printers. set fontlwidth line-width This sets the width of lines to be used to draw the stroked (Plotter fonts) on a PostScript printer. This has a great effect on their appearance. set hei character-size Sets the height of text. For historical reasons, concerning lead type and printing conventions, a height of 10cm actually results in capital letters about 6.5cm tall. set join mitre | round | bevel Defines how two wide lines will be joined together. With mitre, the outside edges of the join are extended to a point and then chopped off at a certain distance from the intersection of the two lines. With round, a curve is drawn between the outside edges. set just left | center | right | tl | etc... Sets the justification which will be used for text commands. set lstyle line-style Sets the current line style to line style number line-style. There are 9 predefined line styles (1--9). When a line style is given with more than one digit the first digit is read as a run length in black, the second a run length in white, the third a run length in black, etc. set lwidth line-width Sets the width of lines to line-width cm. A value of zero will result in the device default of about 0.02 cm, so a lwidth of .0001 gives a thinner line than an lwidth of 0. sub sub-name parameter1 parameter2 etc Defines a subroutine. The end of the subroutine is denoted with end sub. Subroutines must be defined before they are used. Subroutines can be called inside any GLE expression, and can also return values. The parameters of a subroutine become local variables. Subroutines are reentrant. sub tree x y a$ amove x y rline 0 1 write a$ return x/y end sub @tree 2 4 "mytree" (Normal call to subroutine) slope = tree(2,4,"mytree") (Using subroutine in an expression) text unquoted-text-string This is the simplest command for drawing text. The current point is unmodified after the text is drawn so following one text command with another will result in the second line of text being drawn on top of the first. To generate multiple lines of text, use the begin text end text construct. text "Hi, how's tricks", said Jack! write string$ This command is similar to text except that it expects a quoted string, string variable, or string expression as a parameter. The built in functions sqrt() and time$() are described in appendix A.3. commanddescription Expressions ----------- Wherever GLE is expecting a number it can be replaced with an expression. For example rline 3 2 and rline 9/3 sqrt(4) will produce the same result. An expression in GLE is delimited by white space, so it may not contain any spaces - ` rline 3*3 2' is valid but ` rline 3 * 3 2' will not work. Or ` let d2 = 3+sin(d1)' will work and ` let d2= 3 + sin(d1) ' won't. Expressions may contain numbers, arithmetic operators ( +, -, *, /, ^(to the power of)), relational operators ( >, <, =>, <=, =, <>) boolean operators ( AND, OR), variables and built-in functions. When GLE is expecting a color or marker name (like `green' or `circle') it can be given a string variable, or an expression enclosed in braces). GLE provides a large number of built in functions, these are listed in Appendix A.3. Functions Inside Expressions ---------------------------- xg(), yg() With these functions it is possible to move to a position on a graph using the graph's axis units. To draw a filled box on a graph, at position x=948, y=.004 measured on the graph axis: begin graph xaxis min 100 max 2000 yaxis min -.01 max .01 ... end graph amove xg(948) yg(.004) box 2 2 fill grey10 xend(), yend() These functions return the end point of the last thing drawn. This is of particular interest when drawing text. text abc set color blue text def This would draw the def on top of the abc. To draw the def immediately following the abc simply do the following (Note that absolute move is used, not relative move): xpos(), ypos() Returns the current x and y points. commanddescription The Graph Module ---------------- A graph should start with begin graph and end with end graph. The data to be plotted are organised into datasets. A dataset consists of a series of (X,Y) coordinates, and has a name based on the letter ``d'' and a number between 1 and 99, eg. d1 The name dn can be used to define a default for all datasets. Many graph commands described below start with d n. This would normally be replaced by a specific dataset number e.g., d3 marker diamond For each xaxis command there is a corresponding yaxis, y2axis and x2axis command for setting the top left and right hand axes. These commands are not explicitly mentioned in the following descriptions. Graph Commands (a summary) -------------------------- data filename [d1 d2 d3 ... ] [d1=c1,c3 ] dn bigfile "all.dat,xc,yc" [marker mname] [line] dn err d5 errwidth width-exp errup nn% errdown d4 dn herr d5 herrwidth width-exp herrleft nn% errright d4 dn key "Dataset title" dn line dn lstyle line-style lwidth line-width color col dn marker marker-name [msize marker-size] [mdata dn] dn nomiss dn smooth | smoothm dn xmin x-low xmax x-high ymin y-low ymax y-high fullsize hscale exp key pos tl nobox hei exp offset xexp yexp let ds = exp [from low to high step exp ] nobox size x y title "title" [hei ch-hei] [color col] [font font] [dist cm] vscale exp x2labels on xaxis | yaxis | x2axis | y2axis xaxis color col font font-name hei exp-cm lwidth exp-cm xaxis dsubticks sub-distance xaxis grid xaxis log xaxis min low max high dpoints n xaxis nofirst nolast xaxis nticks number dticks distance xaxis off xaxis shift cm-exp xlabels font font-name hei char-hei color col xnames "name" "name" ... xplaces pos1 pos2 pos3 ... xside color col lwidth line-width off xsubticks lstyle num lwidth exp length exp off xticks lstyle num lwidth exp length exp off xtitle "title" [hei ch-hei] [color col] [font font] [dist cm] y2title "text-string" [rotate] bar dx,... dist spacing bar dx,... from dy,... bar dn,... width xunits,... fill col,... color col,... fill x1, d3 color green xmin val xmax val fill d4,x2 color blue ymin val ymax val fill d3,d4 color green xmin val xmax val fill d4 color green xmin val xmax val Graph Commands (in detail) -------------------------- data filename [d1 d2 d3 ... ] [d1=c1,c3 ] Specifies the name of a file to read data from. By default, the data will be read into the next free datasets unless the optional specific dataset names are specified. A dataset consists of a series of (X,Y) coordinates, and has a name based on the letter d and a number between 1 and 99, e.g. d1 or d4. Up to 99 datasets may be defined. From a file with 3 columns the command ` data xx.dat' would read the first and second columns as the x and y values for dataset 1 (d1) and the first and third columns as the x and y values for dataset 2 (d2). The next data command would use dataset 3 (d3). A data file for two datasets looks like this (*=missing value): 1 2.7 3 2 5 * 3 7.8 7 4 9 4 The first coordinate of dataset d1 would then be ( 1,2.7) and the first coordinate of dataset d2 would be ( 1,3). The option d3=c2,c3 allows particular columns of data to be read into a dataset, d3 would read x values from column 2 and y values from column 3. dn bigfile "all.dat,xc,yc" [marker mname] [line] The bigfile option allows a dataset to be read as it is drawn, (rather than being complete read into memory before it is drawn) this means that very large datasets can be drawn on a PC without running out of memory. The axis minimum and maximum must be specified (using the command xaxis min exp max exp. By default the first two columns of the data file will be read in, but other columns may be specified. E.g., all.dat,3,2 would read x values from column 3 and y values from column 2. Or, to read the 4th dataset, specify the file as all.dat,1,5 If the x column is specified as '0' then GLE will generate the x data points. E.g., 1,2,3,4,5... Many (but not all) of the normal dn commands can be used with the bigfile command. E.g., marker, lstyle, xmin, xmax, ymin, ymax, color and lwidth. You cannot use commands like let or bar with the bigfile command. dn err d5 errwidth width-exp dn errup nn% errdown d4 For drawing error bars on a graph. The error bars can be specified as an absolute value, as a percentage of the y value, or as a dataset. The up and down error bars can be specified separately e.g., d3 err .1 d3 err 10% d3 errup 10% errdown d2 d3 err d1 errwidth .2 dn herr d5 herrwidth width-exp dn herrleft nn% errright d4 These commands are identical to the error bar commands above except that they will draw bars in the horizontal plane. dn key "Dataset title" If a dataset is given a title like this a key will be drawn. Use the key command (below, after hscale) to set the size and position of the key. Use the key module (Chapter 4) to draw more complex keys. dn line This tells GLE to draw lines between the points of the dataset. By default GLE will not draw lines or markers, this is often the reason for a blank graph. If a dataset has missing values GLE will not draw a line to the next real value, which leaves a gap in the curve. To avoid this behavior simply use the nomiss qualifier on the dn command used to define the line. This simply throws away missing values so that lines are drawn from the last real value to the next real value. dn lstyle line-style lwidth line-width color col These qualifiers are all fairly self explanatory. See the lstyle command in Chapter~2 for details of specifying line styles. dn marker marker-name [msize marker-size] [mdata dn] Specifies the marker to be used for the dataset. There is a set of pre-defined markers (refer to Appendix A.1 for a list) which can be specified by name (e.g., circle, square, triangle, diamond, cross, ...). Markers can also be a user-defined subroutine (See the define marker command in Chapter 2). The mdata option allows a secondary dataset to be defined which will be used to pass another parameter to the marker subroutine, this allows each marker to be drawn at a different angle,size or color. The msize qualifier sets the marker size for that dataset. The size is a character height in cm, so that the actual size of the markers will be about 0.7 of this value. dn nomiss If a dataset has missing values, GLE will not draw a line to the next real value, which leaves a gap in the curve. To avoid this behavior simply use the nomiss qualifier on the dn command used to define the line. This simply ignores missing values. dn smooth | smoothm This will make GLE draw a smoothed line through the points. A third degree polynomial is fitted piecewise to the given points. The smoothm alternative will work for multi valued functions, i.e., functions which have more than one y value for each x value. dn xmin x-low xmax x-high ymin y-low ymax y-high These commands map the dataset onto the graph's boundaries. The data will be drawn as if the X axis was labelled from x-low to x-high (regardless of how the axis is actually labelled). A point in the dataset at X = x-low will appear on the left hand edge of the graph. fullsize This is equivalent to vscale 1, hscale 1, noborder. It makes the graph size command specify the size and position of the axes instead of the size of the outside border. hscale exp Scales the length of the yaxis. See vscale. The default value is 0.7. key pos tl nobox hei exp offset xexp yexp This command allows the features of a key to be specified. The pos qualifier sets the position of the key. E.g., tl=topleft, br=bottomright, etc. let ds = exp [ from low to high step exp ] This command defines a new dataset as the result of an expression on the variable x over a range of values. It also allows the use of other datasets. E.g., to generate an average of two datasets: data aa.dat d1 d2 let d3 = d1+d2/2 Or to generate data from scratch: let d1 = sin(x)+log(x) from 1 to 100 step 1 If the xaxis is a LOG axis then the step option is read as the number of steps to produce rather than the size of each step. NOTE: The spacing around the `=' sign and the lack of spaces inside the expression are necessary. nobox This removes the outer border from the graph. size x y Defines the size of the graph in cm. This is the size of the outside box of a graph. The default size of the axes of the graph will be 70% of this, (see vscale and hscale). This command is required. title "title" [hei ch-hei] [color col] [font font] [dist cm] This command gives the graph a centred title. The list of optional keywords specifies features of it. The dist command is used for moving the title up or down. vscale exp This sets the width of the axis relative to the width of the graph. For example with a 10cm wide graph and a vscale of .6 the x axis would be 6cm long. A setting of 1.0 makes the xaxis the same length as the width of the graph, which is useful for positioning some graphs. The default value is 0.7. x2labels on This command `activates' the numbering of the x2axis. There is a corresponding command ` y2axis on' which will activate y2axis numbering. xaxis | yaxis | x2axis | y2axis A graph is considered to have four axes: The normal xaxis and yaxis as well as the top axis (x2axis) and the right axis (y2axis). Any command defining an xaxis setting will also define that setting for the x2axis. The secondary axes x2 and y2 can be modified individually by starting the axis command with the name of that axis. E.g., xaxis color col font font-name hei exp-cm lwidth exp-cm These axis qualifiers affect the color, lstyle, lwidth, and font used for drawing the xaxis (and the x2axis). These can be overriden with more specific commands. E.g., ` xticks color blue' would override the axis color when drawing the ticks. The subticks would also be blue as they pick up tick settings by default. xaxis dsubticks sub-distance See xaxis nticks below. xaxis grid This command makes the xaxis ticks long enough to reach the x2axis and the yaxis ticks long enough to reach the y2axis. When used with both the x and y axes this produces a grid over the graph. Use the xticks lstyle command to create a faint grid. xaxis log Draws the axis in logarithmic style, and scales the data logarithmically to match (on the x2axis or y2axis it does not affect the data, only the way the ticks and labelling are drawn) Be aware that a straight line should become curved when drawn on a log graph. This will only happen if you have enough points or have used the smooth option. xaxis min low max high dpoints n Sets the minimum and maximum values on the xaxis. This will determine both the labelling of the axis and the default mapping of data onto the graph. To change the mapping see the dataset dn commands xmin, ymin, xmax, and ymax. The dpoints option specifies the number of decimal points that should be displayed. e.g. with a value of 3 these numbers (0, 0.333333, 0.6666667, 1) would become (0.000, 0.333, 0.667, 1.000) xaxis nofirst nolast These two switches simply remove the first or last (or both) labels from the graph. This is useful when the first labels on the x and y axis are too close to each other. xaxis nticks number dticks distance nticks specifies the number of ticks along the axis. dticks specifies the distance between ticks and dsubticks specifies the distance between subticks. For example, to get one subtick between every main tick with main ticks 3cm apart, simply specify dsubticks 1.5. By default ticks are drawn on the inside of the graph. To draw them on the outside use the command: xticks length -.2 yticks length -.2 xaxis off Turns the whole axis off --- labels, ticks, subticks and line. Often the x2axis and y2axis are not required, they could be turned off with the following commands: x2axis off y2axis off xaxis shift cm-exp This moves the labelling to the left or right, which is useful when the label refers to the data between the two values. xlabels font font-name hei char-hei color col This command controls the appearance of the axis labels but not the axis title. xnames "name" "name" ... This command replaces the numeric labelling with absolutely anything. Given data consisting of five measurements, taken from Monday to Friday, one per day then xaxis min 0 max 6 dticks 1 xnames "" "Mon" "Tue" "Wed" "Thu" "Fri" "" would give the desired result. Note it is essential to define a specific axis minimum, maximum, dticks, etc., otherwise the labels may not correspond to the data. If there isn't enough room on the line for all the names then simply use an extra xnames command. xplaces pos1 pos2 pos3 ... This is similar to the xnames command but it specifies a list of points which should be labelled. This allows labelling which isn't equally spaced. The above example with days of the week could be written like this: xplaces 1 2 5 xplaces 7 xnames "Mon" "Tue" "Fri" "Sun" If there isn't enough room on the line for all the names then simply use an extra xplaces command. xside color col lwidth line-width off This command controls the appearance of the axis line, i.e. the line to which the ticks are attached. xsubticks lstyle num lwidth exp length exp off This command gives fine control of the appearance of the axis subticks. xticks lstyle num lwidth exp length exp off This command gives fine control of the appearance of the axis ticks. Note: To get ticks on the outside of the graph, i.e. pointing outwards, specify a negative tick length: xticks length -.2 yticks length -.2 xtitle "title" [hei ch-hei] [color col] [font font] [dist cm] This command gives the axis a centered title. The list of optional keywords specify features of it. The dist command is used for moving the title up or down. y2title "text-string" [rotate] By default the y2title is written vertically upwards. The optional rotate keyword changes this direction to downwards. The rotate option is specific to the y2title command. commanddescription Bar Graphs ---------- Drawing a bar graph is a subcommand of the normal graph module. This allows bar and line graphs to be mixed. The bar command is quite complex as it allows a great deal of flexibility. The same command allows stacked, overlapping and grouped bars. For stacked bars use separate bar commands as in the first example below: bar d1 fill black bar d2 from d1 fill grey10 For grouped bars put all the datasets in a list on a single bar command: bar d1,d2,d3 fill grey10,grey40,black bar dx,... dist spacing Specifies the distance between bars in dataset(s) dx,.... The distance is measured from the left hand side of one bar to the left hand side of the next bar. A distance of less than the width of a bar results in the bars overlapping. bar dx,... from dy,... This sets the starting point of each bar in datasets dx,... to be at the value in datasets dy,..., and is used for creating stacked bar charts. Each layer of the bar chart is created with an additional bar command. bar d1,d2 bar d3,d4 from d1,d2 bar d5,d6 from d3,d4 Note 1: It is important that the values in d3 and d4 are greater than the values in d1 and d2. Note 2: Data files for stacked bar graphs should not have missing values, replace the * character with the number on its left in the data file. bar dn,... width xunits,... fill col,... color col,... The rest of the bar qualifiers are fairly self explanatory. When several datasets are specified, separate them with commas (with no spaces between commas). bar d1,d2 width 0.2 dist 0.2 fill grey10,grey20 color red,green commanddescription Filling Between Lines --------------------- fill x1, d3 color green xmin val xmax val Fills between the xaxis and a dataset, use the optional xmin, xmax, ymin, ymax qualifiers to clip the filling to a smaller region fill d4,x2 color blue ymin val ymax val This command fills from a dataset to the x2axis. fill d3,d4 color green xmin val xmax val This command fills between two datasets. fill d4 color green xmin val xmax val This command treats the dataset as a polygon and fills it. The dataset should be a closed polygon. commanddescription Notes on Drawing Graphs ----------------------- Importance of Order Most of the graph commands can appear in any order, but in some cases order is significant. As some let commands operate on data which has been read into datasets, the data commands should precede the let commands. The wildcard dn command should appear before specific d1 commands which it will override. By default xaxis commands also change the x2axis, and xlabels commands also change x2labels, so to specify different settings for the x and x2 axes, put the x2 settings after the x settings. begin graph size 10 10 data a.dat let d2 = d1*3 dn marker square lstyle 3 ! sets d1 and d2 d2 marker dot xaxis color green xticks color blue x2axis color black end graph Line Width When scaling a graph up or down for publication the default line width may need changing. To do this simply specify a set lwidth command before beginning the graph. size 10 10 set lwidth .1 begin graph ... end graph The KEY module -------------- This module is used for drawing keys for graphs. It is completely separate from the graph module. Specify a position for the key, the size of the lettering and then for each dataset specify marker, color, lstyle and fill pattern: begin graph ... end graph begin key hei .5 position tr text "Green trees" marker heart msize .2 fill grey10 text "Red trees" lstyle 3 marker heart fill blue text "Oranges" marker circle lstyle 6 fill black end key All commands to do with a particular line of the key module MUST appear on the same line. Key commands ------------ offset x-exp y-exp Specifies the offset in cm from the current point to the bottom left hand corner of the graph. This command should be on a line of its own. position justify-exp This is an alternative to the OFFSET command. It allows you to specify a position on the graph, e.g., tl = top left. This command should be on a line of its own. See set just for a list of justify settings. text str-exp The text which will be displayed on the end of the line. lstyle style-num The line style which will be used for the short line drawn in the key. marker marker-name The marker which will be used for that line of the key. msize exp Specifies the size of the markers in cm. mscale exp Specifies how much to scale the size of the marker. E.g., 0.5 would produce a marker half as big as normal. color color-name The colour of the text, line and marker. hei cm-exp This sets the height of the text used to draw the key. The key will change in size to fit around the text. If you omit this command the current font size is used. fill fill-pattern The fill pattern used in that line of the key. commanddescription Advanced features of GLE ------------------------ This chapter covers the advanced features of GLE. Some of these features will not give exactly the same results on the PC screen as you would get when printing to a PostScript printer. E.g., clipping to an arbitrary shape is only implemented in the PostScript driver but most other features will give as close as possible representation given the limitations of the screen. Color ----- Internally GLE treats color and fill identically, they are simply an intensity of Red, Green and Blue. Each of the predefined color names (yellow,grey20,orange,red) simply define the ratio of red, green and blue. There are two ways to use variables to show color, one is for shades of grey: for i = 0 to 10 box 3 .2 fill (i/10) rmove 0 .2 next i The other is for passing a color name as a variable: sub stick c$ box .2 2 fill c$ end sub @stick "green" Remember a fill pattern completely obscures what is behind it, so the following command would produce a box with a shadow: amove 4 4 box 3 2 fill grey10 rmove -.1 .1 box 3 2 fill white rmove .4 .4 text hellow Device Drivers -------------- GLE supports the following devices. Interactive: IBM/PC (BGI), VT100, REGIS (VT125, VT240), TEK4010, VWS, XWindows. Output: PostScript, HPGL, EPSON, EPSON 24pin, HP Deskjet. Keyboard Mappings: VT100 VT200 PC Meaning GOLD 1 F11 F1 Help GOLD 2 F12 F2 Save GOLD 3 F13 F3 Load GOLD 4 F14 F4 Save-as GOLD 9 F9 F9 Graph-menu GOLD 0 F10 F10 Draw-it Control+Z Control+Z Control+Z Exit/Escape Alt+X Exit/Escape Control+E Control+E Calls VAX EDT Control+F Control+F Toggle fast/slow text Control+R Control+R F5 Shows errors Control+S Shells to DOS PC Screen Drivers Remember that what you see on the screen isn't always what you will get on the printer. For example filled regions will not be filled, and some characters may not look right. After pressing F10 and drawing the graph it can be annotated by using the mouse (or arrow keys) to draw lines, text and boxes. To draw lines simply click on the points of the line, use the right hand mouse button to `pick up' the pen. To draw text press the letter `t' and then click on where you would like the text to be drawn. All movements are rounded to the grid size settings which are 1.0cm, 0.1cm, 0.01cm etc. The height and color of the text/lines is determined by the current settings at the end of the GLE file. If there is no mouse driver loaded then a cross-hair will appear and it can be moved around using the arrow keys. Press `c' to click, instead of the mouse button. PostScript Driver To print a GLE file to the laser printer type: $ cgle myfile /print or on a PC: C:\GLE> psgle myfile C:\GLE> print myfile.ps The postscript drivers for GLE will automatically flip a picture to best fit onto the page, e.g. a wide graph (as defined by the size command at the top) will be drawn in landscape mode and a tall thin graph will be drawn in portrait mode. To produce an .eps file on a VAX for inclusion in LaTeXor WordPerfect you would type: $ cgle myfile /dev=eps On a PC you would type: C:\GLE> psgle myfile /eps (this creates myfile.eps, not myfile.ps) Inside your LaTeXdocument use the LaTeXcommand: \graphin{myfile.eps}{12.0cm}{3.0cm}{1.0} The width and height are used by LaTeXto reserve enough space for the drawing and the last parameter is a scale factor. The laser printer driver will draw all zero width lines .02cm wide for any line width equal to zero, but if an lwidth is greater than zero and less than or equal to .0001 then it will use a line width of 1 pixel. Without this it would be impossible to specify a line width that didn't occasionally get rounded to 2 pixels. TEK4010 Driver This driver allows initialization sequences to be defined with the symbols TEK_OPEN and TEK_CLOSE. On a VAX this is normally done by CGLECMD.COM so that when you specify /DEV=V550 the assignments are done for you. (V550 = Visual 550) HPGL Driver Device Drivers>HPGL Driver This driver allows initialization sequences to be defined in the symbols HPGL_OPEN and HPGL_CLOSE. On the PC use environment variables and on the VAX use DCL symbols. Also HPGL_WIDTH and HPGL_HEIGHT can be defined for non A3 plotters. On a VAX this is normally done by CGLECMD.COM so that when you specify /DEV=HPA4 the assignments are done for you. The HPGL driver assigns the following colors to pen numbers: 1=black, 2=red, 3=green, 4=blue, 5=magenta, 6=white PC Bitmap Drivers GLE supports the EPSON 8 and 24 pin and HP deskjet/laserjet printers. To support bitmap devices which require a large amount of memory GLE first writes a device independent file OUT.DVI, then the appropriate bitmap driver for your printer will read the OUT.DVI file and create a bitmap which it then prints to LPT1: C:> dvigle myfile (produces OUT.DVI) C:> dviepson (creates bitmap and prints to LPT1:) The output options are: C:> dviepson Standard EPSON printers C:> dviep24 24 Pin EPSON printers (180dpi) C:> dvilj HP Laser jet, Desk jet (150 dpi) C:> dvilj300 HP Laser jet, Desk jet (300 dpi) The high resolution drivers (dviep24, dvilj300) are significantly slower than the low resolution drivers so would only be used for final output. See the file AAREADME.GLE for the latest information on drivers. Fonts (font mapping) By default the generic fonts (rm, rmb, ss, tt etc) will all map to PLSR (plotter simplex roman) on BITMAP and HPGL drivers. To make this happen on other drivers put the command plotter fonts immediately after the size command at the top of the GLE file. A typical result of this change in fonts is that something that lines up on the screen will not line up when printed to an EPSON printer. If this happens then use the plotter fonts command. If a character is missing from a font, or isn't the particular variation you like, you can define a character to be from a different font in this way: \chardef{%}{{\setfont{texcmr}\char{37}}} \chardef{[}{{\setfont{texcmr}\char{91}}} \chardef{]}{{\setfont{texcmr}\char{93}}} On the PC some fonts may not be installed to save disk space. When one of the missing fonts is called for, a replacement font will be displayed, this may look terrible and some special characters may be completely wrong. More importantly if you use a font which does not have its font metric file installed (e.g. C:/GLE/FONTS/PLSR.FMT) then the PostScript driver will space the characters incorrectly. This can be fixed by extracting that particular metric file from the distribution file CGLE_FVE.ZIP using PKUNZIP. Diagrams, Joining Named Objects ------------------------------- To draw lines between boxes which contain text, first name each box as it is drawn and then use the join command to draw the lines between the boxes. box 2 3 fill blue name square amove 5 5 begin box add .1 name titlebox text Title end box join square.tr -> titlebox.bc These commands draw a line from the ``Top Right'' of the square to the ``Bottom Centre'' of the titlebox, with an arrow at the titlebox end. join square - titlebox would draw a line from the centre of the square to the centre of the titlebox but clipped correctly at the edges of both boxes. join square.tc <-> titlebox.v would draw a vertical line from the top centre of the square to the titlebox with arrows at both ends. Named points on each box: .bl Bottom left .bc Bottom centre .br Bottom right .cr Centre right .tr Top right .tc Top centre .tl Top left .cl Centre left .v Vertical line .h Horizontal line .cc Centre centre .ci Circle clipping (for drawing lines to a circle) To draw lines to a given point, simply move there and save that point as a named object. rmove 2 3 save apoint join apoint - square Filling, Stroking and Clipping Paths ------------------------------------ It is possible to set up arbitrary clipping regions. To do this draw a shape and make it into a path by putting a begin path clip ... end path, around it. Then draw the things to be clipped by that region. To clear a clipping path surround the whole section of GLE commands with begin clip ... end clip Characters can be used to make up clipping paths, but only the PostScript fonts will currently work for this purpose. Clipping doesn't work on the screen or p79 devices, but does on the laser printer. See example CLIP.GLE at end of manual. size 10 5 begin clip ! Save current clipping path begin path clip stroke ! Define new clipping region amove 2 2 box 3 3 amove 6 2 box 3 3 end path amove 2 2 set hei 3 text Here is clipped text end clip ! Restore original clipping path Using Variables --------------- GLE has two types of variables, floating point and string. String variables always end with a dollar sign. A string variable contains text like ``Hello this is text'', a floating point variable can only contain numbers like 1234.234. name$ = "Joe" height = 6.5 ! Height of person shoe = .05 ! shoe adds to height of person amove 1 1 box .2 height+shoe write name$ Programming Loops ----------------- The simple way to draw a 6 $$ 8 grid would be to use a whole mass of line commands: amove 0 0 rline 0 8 amove 1 0 rline 1 8 ... amove 6 0 rline 6 8 this would be laborious to type in, and would become impossible to manage with several grids. By using a simple loop this can be avoided: for x = 0 to 6 amove x 0 rline x 8 next x for y = 0 to 8 amove 0 y rline 6 y next y To draw lots of grids all of different dimensions a subroutine can be defined and then used again and again: ! define the subroutine sub grid nx ny gsave begin origin for x = 0 to nx amove x 0 aline x ny next x for y = 0 to ny amove 0 y aline nx y next y end origin end sub ! now draw the grids wherever amove 2 4 @grid 6 8 amove 2 2 @grid 9 5 Now the main GLE file will be much easier to modify particularly if the subroutine definition is moved into a separate file: size 10 10 include griddef.gle amove 2 4 @grid 2 4 amove 2 2 @grid 9 5 Tables ------ Markers ------- Fonts ----- font-name Description rm Roman rmb Roman Bold rmi Roman Italic ss San Serif ssb San Serif Bold ssi San Serif Italic tt Typewriter ttb Typewriter Bold tti Typewriter Italic font-name Description psagb AvantGarde-Book psagbo AvantGarde-BookOblique psbd Bookman-Demi psbdi Bookman-DemiItalic psbl Bookman-Light psbli Bookman-LightItalic psc Courier pscb Courier-Bold pscbo Courier-BoldOblique psco Courier-Oblique psh Helvetica pshb Helvetica-Bold pshbo Helvetica-BoldOblique psho Helvetica-Oblique psncsb NewCenturySchlbk-Bold psncsi NewCenturySchlbk-Italic psncsr NewCenturySchlbk-Roman pspr Palatino-Roman pstr Times-Roman psti Times-Italic pstb Times-Bold pstbi Times-BoldItalic pszcmi ZapfChancery-MediumItalic pszd ZapfDingbats pssym Symbol font-name Description texcmb Computer Modern Bold texcmex Computer Modern Extensible texcmitt Computer Modern Italic Typewriter texcmmi Computer Modern Maths Italic texcmr Computer Modern Roman texcmss Computer Modern Sans Serif texcmssb Computer Modern Sans Serif Bold texcmssi Computer Modern Sans Serif Italic texcmsy Computer Modern Symbol texcmti Computer Modern Text Italic texcmtt Computer Modern Typewriter Text font-name Description plcc Complex Cartographic plcg Complex Gothic plci Complex Italic plcr Complex Roman plcs Complex Script pldr Duplex Roman plge Gothic English plgg Gothic German plgi Gothic Italian plsa Simplex Ascii plsg Simplex German plsr Simplex Roman plss Simplex Script plsym1 Symbols one plsym2 Symbols two plti Triplex Italic pltr Triplex Roman plba Block Ascii Functions --------- Function Name Returns TIME$() current time e.g. ``11:44:27'' DATE$() current date e.g. ``Tue Apr 09 1991'' LEFT$(str$,exp) left exp characters of str$ RIGHT$(str$,exp) rest of str$ starting at exp SEG$(str$,exp1,exp2) str$ from exp1 to exp2 NUM$(exp) string representation of exp NUM1$(exp) as above but with no spaces VAL(str$) value of the string str$ POS(str1$,str2$,exp) position of str2$ in str1$ from exp LEN(str$) the length of str$ Function Name Returns ABS(exp) absolute value of expression ATN(exp) arctan COS(exp) cosine EXP(exp) exponent FIX(exp) exp rounded towards 0 INT(exp) integer part of exp LOG(exp) log to base e of exp LOG10(exp) log to base 10 of exp SGN(exp) returns 1 or -1. SIN(exp) sine of exp SQR(exp) exp squared TAN(exp) tangent of exp NOT(exp) logical not of exp RND(exp) random number from seed exp SQRT(exp) square root of exp Function Name Returns XEND() the x end point of a text string when drawn YEND() the y end point of a text string when drawn XPOS() the current x point YPOS() the current y point TWIDTH(str$) the width of str$ assuming current font, size THEIGHT(str$) the height of str$ assuming current font, size TDEPTH(str$) the depth of str$ XG(xexp) converts units of last graph to abs cm. YG(yexp) converts units of last graph to abs cm. tabular