Kaleidagraph revision: your suggestions invited
Jeff Bailey
Bill Jones <mailto:jo...@synergy.com> at Synergy Software
<http://www.synergy.com> tells me that a revision of KaleidaGraph for the
Mac is in the works, and is inviting user suggestions for what we'd like
to see in the next version. I've already sent him mine, and he would like
to see suggestions from other users as well. Development for the Mac has
been stalled while the Windoze version of KGraph was cranked out, but now
it's time for a new Mac version.
When you reply, please be as specific as possible and focus on possible
improvements to KaleidaGraph. (Please don't use this opportunity to vent
about the relative merits of the Mac or other platforms--he says this
happened the last time he invited suggestions.)
About Kaleidagraph:
"KaleidaGraph is an award-winning graphing solution since 1988, abundant
with features, yet kind to the budget. Easy data importing, versatility in
graphing styles, and control of all plot elements allow you to create
meaningful graphs. Curve fitting and error bars help complete an
insightful overall analysis. The current version of KaleidaGraph for
Macintosh is 3.0.5." Check out more details at <http://www.synergy.com>.
Thanks, Jeff
--
Jeff Bailey
rjba...@scruznet.com
http://www.scruz.net/~rjbailey/
Jie Yuan
In article <rjbailey-200...@rjbailey.sc.scruznet.com>,
rjba...@scruznet.com (Jeff Bailey) wrote:
> Bill Jones <mailto:jo...@synergy.com> at Synergy Software
> <http://www.synergy.com> tells me that a revision of KaleidaGraph for the
> Mac is in the works, and is inviting user suggestions for what we'd like
> to see in the next version. I've already sent him mine, and he would like
> to see suggestions from other users as well. Development for the Mac has
> been stalled while the Windoze version of KGraph was cranked out, but now
> it's time for a new Mac version.
I have a few:
1. Ability to keep show-page info in a graphic file so that the same
layout can be generated next time this file is opened. This one is
STRONGLY desired, and very urgent.
2. Ability to export GIF, JPG, TIFF graphics, in addition to PICT. This
one is not urgent.
3. Multi-dimentional plots and curve fits (not urgent).
4. Ability to select multiple macro files (libraries) and open them upon
start. When a custome macro is not loaded when called, prompt for loading
the library that contains it (not urgent).
5. Be savvy with VM, RAM-Doubler, SpeedDoubler, PPC, and common extensions.
Guess I stop here. I may post some more tomorrow.
--
-- Jie Yuan - Dept. of Pharmacology & Cell Biophysics - U. Cin. --
-- Cin, OH 45267-0575 - Fax: 513-558-1169 - using NewsWatcher --
James Bridges
In article <rjbailey-200...@rjbailey.sc.scruznet.com>
rjba...@scruznet.com (Jeff Bailey) writes:
> When you reply, please be as specific as possible and focus on possible
> improvements to KaleidaGraph. (Please don't use this opportunity to vent
> about the relative merits of the Mac or other platforms--he says this
> happened the last time he invited suggestions.)
>
One outstanding (for me) feature of the plot templates (the example
plots used when running the Plot Script) is that Plot Script uses all
of the same settings except the axes ranges--it always autoranges.
There should be a way to fix these ranges so that every plot created by
Plot Script has the same range. I also often have problems with Plot
Script while creating several multi-file plots. I get the files plotted
in the wrong order, spend time fixing it in the Data Windows of the
Plot Script and often have the program ignore my changes unless I get
the correct file window in front of the others. Very confusing if
nothing else.
The only new features I could ask for is 3D, but I'd hate to see the
product get put in a higher price bracket for this. Another thing I'd
hate to see would be "wizards" a la Excel and DeltaGraph, and other
business-type enhancements. KG is a good technical plotting package; I
hope it can survive as such.
KaleidaGraph is so good as is it would be easy to screw it up.
James Bridges
Cleveland, OH, USA
By day: aeroacoustics researcher at NASA Lewis Research Center
By night: mad scientist
Quote: Turbulence is a point of view.
Michael B. Moore
In Article <rjbailey-200...@rjbailey.sc.scruznet.com>,
rjba...@scruznet.com (Jeff Bailey) wrote:
>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>Mac is in the works, and is inviting user suggestions for what we'd like
>to see in the next version. I've already sent him mine, and he would like
>to see suggestions from other users as well. Development for the Mac has
>been stalled while the Windoze version of KGraph was cranked out, but now
>it's time for a new Mac version.
>
1. 3D Plotting, including xyz coordinate (floating-point in all 3 directions).
2. More than 9 plot limit per plot (Say, 20?)
3. Direct pasting of coordinates of points into a data window using the
Identify Tool.
4. Ability to copy multiple column titles into the clipboard from the Column
Format dialog, i.e., ability to copy more than one title into the clipboard
without copying the associated data.
Craig Jones
In Article <rjbailey-200...@rjbailey.sc.scruznet.com>,
rjba...@scruznet.com (Jeff Bailey) wrote:
>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>Mac is in the works, and is inviting user suggestions for what we'd like
>to see in the next version.
(1) Text editing on the page, not in a funky dialog box. Regular access to
fonts.
(2) Be able to change some of the characteristics of a bunch of plotted
points/lines at one swoop (e.g., shift-click to select several points in the
legend, then use the pop-up on the tool palette to resize all the symbols).
(3) Symbols really any size (even though I can enter any number in the Plot
Styles dialog, Kgraph seems to always go to one of its fixed collection of
sizes. Most annoying when I want size 6 and get 4).
(4) multidimensional curve fitting
(5) assign symbol size/color/shape by third/fourth/fifth variable. (this I
would REALLY REALLY like--I have to hack a bunch of new columns right now
using if statements and then get everything straight when plotting).
(6) fixed aspect ratio (especially useful if making plots for a map base)
(7) binary data input complete with byte offset and byte swapping ability
(8) recordable AppleScripting sounds nice
(9) a dream, perhaps, but I've never had much luck with the programmable
calculator, though I do ok with the macro library. A less cryptic
"programming" interface (sort of expanded macros) would be really nice.
(10) histogram bin widths an option instead of simply number of bins.
(11) transpose rows and columns option to allow a first column to go into the
variable names (in fact, transpose could simply just do this whenever the first
column is text).
(12) Cleaner export to PICT of multiple fonts and super/subscripts in text
items. Right now these come through as a bunch of separate items and editing
is very difficult in other programs; also, sub- and super-scripts are frequently
poorly placed in export (though this also sometimes seems to depend on importing
code). Seems like several vendors are making these multifont, multibaseline
text pieces come through to other programs ok. Same with line styles--if I
export dashed lines, they can't be changed to some other linestyle in a
drawing program like Canvas. Again, seems like linestyles are now portable
across programs).
(13) Don't disconnect old plots from current data when a new plot is made.
This drives me nuts--I'll make one plot, say x vs z, then one y vs z, decide
some points should be dropped and so I mask them, but the x vs z plot won't use
the data now and cannot be updated to remove the masked points. I have to
either make a NEW x-z plot (making the y-z plot "old") or extract the
"old" data and mask it the same way as the "new" data. Even if there is some
workaround in the existing version that is not so severe, it isn't particularly
obvious.
(14) THREE error bars/variable (x, y, and label). Yes, I can work around this
right now by plotting my variable twice using X2, but extracting the label
to its own error bar would also permit numerical data to be displayed as
text next to points (this is occasionally useful when there is a third
dimension floating about, and the workaround here--making the variable into
text, or making a new variable that is text, is trouble if you recalculate
anything, etc.). A special label "error bar" would also permit a cleaner
interface to locating where
labels should be plotted relative to the point.
(15) Palette/dialog or even dragging within legend to reorder plotting
order. Right now to get variable A under variable B you have to choose A
first in
making the plot--if you have a bunch of variables across a few files, this
can get to be a serious pain.
Craig Jones cjo...@mantle.colorado.edu
Research Associate, CIRES, University of Colorado, Boulder
WWW: http://cires.colorado.edu/people/jones.craig/CHJ_home.html
Bruce E. Goldstein
In article <rjbailey-200...@rjbailey.sc.scruznet.com>,
rjba...@scruznet.com (Jeff Bailey) wrote:
!Bill Jones <mailto:jo...@synergy.com> at Synergy Software
!<http://www.synergy.com> tells me that a revision of KaleidaGraph for the
!Mac is in the works, and is inviting user suggestions for what we'd like
!to see in the next version. I've already sent him mine, and he would like
SNIP
I would like to see first a way of saving the showpage configuration and
somehow tying this to specific plots, so that I could later reproduce
a stacked set of multiple plots on a single page.
Second, I have found that it would be desireable to specify BOTH frame and
axis. Kaleidagraph usually leaves too much white space around the edge.
This is OK when manually cutting and pasting with real scissors, but when
one exports (e.g.,as PICTs) into an electronic manuscript one is forced to
shrink it down so that the white border fits within the column width,
leaving a shrunken figure in the middle. There are work arounds such as
importing into graphics editing programs, but sometimes these fail because
everything is grouped and ungrouping screws up locations and rotations.
Just importing
a Kaleidagraph plot directly into a manuscript usually looks pretty ugly
because of this white space problem.
Ron Shepard
>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>Mac is in the works, and is inviting user suggestions for what we'd like
>to see in the next version. I've already sent him mine, and he would like
>to see suggestions from other users as well. Development for the Mac has
>been stalled while the Windoze version of KGraph was cranked out, but now
>it's time for a new Mac version.
How about contour plots, with color, and with value-based line types (e.g.
solid for positive values, dashed for negative values)?
How about random dot stereograms? I use Mathematica for this now, but it
would be simpler to use KG for all of my scientific graphing. The dot
shape and size should be variable, along with the dot density. The
imaging plane should be movable to "above" the surface (the normal paper
plots) and also to "behind" the surface (for overhead projector
viewgraphs).
How about combination contour and hidden line plots?
How about allowing the text strings in the Legend box to be edited? In
particular, I would like to paste in general mathematical expressions from
Expressionist or MathType (or MS Word).
$.02 -Ron Shepard
alice faber
>>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
>><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>>Mac is in the works, and is inviting user suggestions for what we'd like
>>to see in the next version. I've already sent him mine, and he would like
>>to see suggestions from other users as well. Development for the Mac has
>>been stalled while the Windoze version of KGraph was cranked out, but now
>>it's time for a new Mac version.
Several of the items from my wish list have already been mentioned in
previous responses, so excuse the repetition:
1) More than 9 x-y combinations per plot. 20 would be a minimum. If you
think that no-one could possibly read such a plot, bear in mind that users
make plots for exploratory reasons as well as for publication.
2) Scrap the "alternate" font business, and replace it with a more
standard text editing interface that would allow free selection of fonts.
I have occasion to use both the symbol font (for Greek letters) and
phonetics fonts (for symbols from the International Phonetics Alphabet) in
labels of various sorts, in addition to whatever standard font I'm using.
I'd like to be able to do this easily.
3) Allow for the use of "text" error bars (aka labels) along with regular
x and y error bars. The availability of text error bars has been the
single feature that has kept me with Kaleidagraph when friends and
colleagues have been raving about everything they can do in Deltagraph,
and I'd like to have this added flexibility.
4) Bivariate confidence ellipses, with user specified radii, and the
option to display only the ellipses (possibly with mean included) and not
all the individual data points. It should be possible to display as many
ellipses as there are plotting symbols; for my own research, 11 ellipses
would be an irreducible minimum.
5) Fix the bug in grid and tick display for minor divisions in plots with
logarithmic axis scales. The bug is that no tick or grid is displayed for
the first minor division. For that matter, more sensible automatic axis
limits for logarithmic plots would be desirable. I commonly plot data
where one axis ranges from 300-900 and the other from 900 to 2500. The
latter commonly defaults to a range from 1 to 10,000; I can't just reset
defaults, since the next data set might have a range of 800 to 2650.
AF
Kelly Miller
>>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
>><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>>Mac is in the works, and is inviting user suggestions for what we'd like
>>to see in the next version. I've already sent him mine, and he would like
>>to see suggestions from other users as well. Development for the Mac has
>>been stalled while the Windoze version of KGraph was cranked out, but now
>>it's time for a new Mac version.
1) Ability to add new data columns while formatting data columns.
(Currently, data columns have to be added seperately from formatting
them.) Generally, more flexibility in formatting the data columns would
be great (perhaps some way to do it without having to go to the dialog box
each time).
2) Improve the interface on the macro calculator. It is very hard to
learn to program it -- much harder than learning to program a programmable
calculator. Ideally, I'd like an interface like a programming language
(say fortran or basic or something similar).
3) Simplify the creation and specification of specific custom graph
styles -- allow easier specification and saving of typefaces and sizes,
less confusing specification of the placement of the graph upon the screen
and the printed page, etc.
4) Add more different sorts of symbols -- there are only five different
sorts which are easily distinguishable from one another (circle, square,
diamond, triangle, inverted triangle). I'd like to see more options, like
heagons, pentagons, five sided stars, etc. Also, when exporting a graph
with these symbols, make the symbols into the appropriate polygon or path
rather than a collection of line segments, in order to make editing of the
graphs easier.
All in all, Kaleidagraph is one of my favorite programs -- it is an
excellent tool for data analysis and graphing.
Kelly Miller
mil...@aries.scs.uiuc.edu
EJ Chichilnisky
>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>Mac is in the works, and is inviting user suggestions for what we'd like
>to see in the next version.
1. Make template plotting more flexible. Allow the user to control what
aspects of a template are used with a dialog with a series of checkboxes
like:
€ Use template X-axis min
€ Use template X-axis max
€ Use template X-axis increment
€ Use template X-axis title
€ Use template plot title
etc...
For each item above that is not checked, the current defaults get used
instead. The above preferences would be stored in the KG preferences
file. Plotting by template would then be infinitely more flexible.
Preferences would default to current standards.
2. Allow flexible positioning of the axes. A typical example is when one
wants the x-axis to inteserct the y-axis at a y-position of 0, and
vice-versa ("crosshair" axes). Generally, the axes should slide anywhere
specified by the user. I am sure you get many requests for this one.
3. More flexbile appleevents/applescript support. The user should be
able to specify all the parameters of a plot, not the limited number of
parameters that you allow through current appleevents interface. The
setup now is clunky and relies (and is limited by) the plot script
interface, which itself is very unintuitive and limited. As another user
commented,
(This one is important! It could lead to KG being split into a
front-end app for
direct user interaction and a plot "server" being driven from the
front end by
means of apple events - this is where mac-gnuplot is heading currently)
I agree and have almost discarded KG on several occasions because of the
limited usefulness of the scripting. The future of scientific
graphing/plotting is in scripting, not in ever more pointing and
clicking. The "Plot Script" window could be entirely replaced by a
sensible model based on AppleScript, which would be good, since it is
buggy and has a confusing interface.
4. Allow at least 32 plots per plot window. Current limit seems
pointless and once in a while is very frustrating.
5. Allow simple extraction to clipboard of the parameters of a curve fit
(currently to get these parameters the user has to do a lot of pointless
pointing & clicking).
6. Allow copying of columns of data with titles (that is, if you copy
several columns from one data sheet into another, the titles should come
with the data!).
7. Make the page layout more flexible. Plots and text should be able to
go anywhere. Not as crucial as above.
8. 3D plots would be nice. Not as crucial as above.
--
EJ Chichilnisky
e...@white.stanford.edu
Robert Herrick
One more suggestion: my biggest gripe about Kaleidagraph has been that it
doesn't export clean post-script output (using either the clip-board or
by creating an EPS file). I've seen a detailed discussion of
work-arounds for this, but in my papers, I have had to copy the graph
into a drawing program like Freehand or Canvas, and then trace the data
with a Bezier curve to get a smooth output. The Kaleidagraph output
itself has a jagged, stairstep appearance, that invariably draws the
comment, "I hope the final version of the paper won't look like this"
from my Professor.
- Bob Herrick
Robert Herrick
>In article <rjbailey-200...@rjbailey.sc.scruznet.com>,
>rjba...@scruznet.com (Jeff Bailey) wrote:
>>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
>><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>>Mac is in the works, and is inviting user suggestions for what we'd like
>>to see in the next version. I've already sent him mine, and he would like
>>to see suggestions from other users as well. Development for the Mac has
>>been stalled while the Windoze version of KGraph was cranked out, but now
>>it's time for a new Mac version.
Kaleidagraph is THE data plotting program of choice here at UCSB, so it's
good to hear they're coming out with an update for the Mac. I hope they
preserve its greatest strength, which is the fact you don't need to read
a manual or do a tutorial to figure it out - Kaleidagraph is an example
of how software should be written in that respect.
My specific suggestions are as follows:
1) Addition of a point editing tool. Currently, we get glitches in our
data every time our lock-in amplifier changes ranges. We have to plot
out the data, zoom in on it, then toggle between the graph and data
window, and delete the rows with bad data points by hand, one row at a
time. This can be rather time-consuming.
We also use the tool where you can "lasso" points (thus eliminating
outliers). I would like to be able to permenantly remove these points,
and then re-smooth the remaining data. Some additional graphical tools
for data editing would be much appreciated.
2) 3-D plotting would be a feature most co-workers of mine would use.
One of Kaleidagraph's strength's is its ability to import a large number
of data formats - hopefully this feature could import most common image
formats (BMP, TIFF, and PICT, to start), as well as more straight-forward
row-and-column data. The ability to interpolate between limited points
with spline-fits makes the data easier to follow than just plotting the
small number of polygons that a small data set would ordinarily form.
Something where points in different height bands appear in different
colors is also nice.
3) Spline fitting, and interpolation for two-dimensional data, and the
ability to take data on an irregular grid of points, and create
interpolated data for any dx-step the user selects.
4) Peak-fitting capability (i.e., The ability to extract amplitudes and
center wavelengths of multiple gaussian (or lorentians) which make up the
data.) While we have a number of programs that do this on our PC's (e.g.,
PKFit, Jandel PeakFit), no one in the department has a comparable mac
program. This feature would get a fair amount of use if it was added to
Kaleidagraph.
- Bob Herrick
____________________________________________________________________________
Robert W. Herrick Ph.D. Student, Electrical Engineering
650...@ucsbuxa.ucsb.edu University of California, Santa Barbara
____________________________________________________________________________
Jeff Bailey
In article <shepard-2309...@macrls.tcg.anl.gov>,
she...@tcg.anl.gov (Ron Shepard) wrote:
|How about allowing the text strings in the Legend box to be edited? In
|particular, I would like to paste in general mathematical expressions from
|Expressionist or MathType (or MS Word).
|
|$.02 -Ron Shepard
Ron:
You CAN edit the text strings in the edit box already! Just click on the
text icon in the toolbar and double-click the legend string you want to
edit. Up comes the text edit box, and you can change the string to
whatever you want.
-Jeff
James Bridges
In article <skok-23099...@news.uni-stuttgart.de>
sk...@itw.uni-stuttgart.de (Holger Skok) writes:
> Seriously, I script a lot with both MacPERL and Applescript and it is a
> bit of a pain
> to piece the plot script together as a single string as required by the
> current scripting
> model of Kaleidagraph. It is a very useful feature, predating Applescript,
> probably, but
> it would be somewhat easier to script KG if the scripting model was
> consistent with
> the Applescript way of doing things: Expose the objects and methods your
> app knows
> about to the world, so other apps and Applescripts can get at them and use them.
> Recordability would be VERY helpful, too.
Yeah! Yeah! I almost gave up scripting because it was too difficult.
Which is faster, writing the AppleScript/KGscript to plot 730 data sets
or doing them by hand? If not for human error due to boredom,
hand-plotting would have been faster.
My vote for true AppleScript support and recordability. Actually, I
think making KG a LiveObject would be great. Think of all the Excel
users that would pick it up to replace the disgusting Graph tool?
Dr N.C. Eastmond
: The only new features I could ask for is 3D, but I'd hate to see the
: product get put in a higher price bracket for this. Another thing I'd
: hate to see would be "wizards" a la Excel and DeltaGraph, and other
: business-type enhancements. KG is a good technical plotting package; I
: hope it can survive as such.
Er no. I don't actually have Kaleidagraph, but I am attracted to it as a
technical graphing utility. I am sick to death of the Deltagraph style
business charting. Wizards can get lost too. All you need is a good graph
format facility (like CA Cricket Graph). No, you can definitely take your
DeltaGraphs and jump. Kaleidagraph sounds just fine as it is. Perhaps the
only thing I might incorporate would be the ablity to perform slightly more
sophisticated stats (without going too far) like ANOVA, t-tests, Mann-
Whitney and maybe Chi-squared.
: KaleidaGraph is so good as is it would be easy to screw it up.
Yep.
--
Nigel C. Eastmond - n...@liv.ac.uk | "My mate Ken can drink this
<http://www.liv.ac.uk/~nce/personal/index.htm> | stuff." - Alas Smith and
Dr.E. - The Cartoon | Jones.
<http://www.liv.ac.uk/~nce/home.htm> |
M. K. Raghuraman
> Bill Jones <mailto:jo...@synergy.com> at Synergy Software
> <http://www.synergy.com> tells me that a revision of KaleidaGraph for the
> Mac is in the works, and is inviting user suggestions for what we'd like
> to see in the next version. I've already sent him mine, and he would like
> to see suggestions from other users as well. Development for the Mac has
> been stalled while the Windoze version of KGraph was cranked out, but now
> it's time for a new Mac version.
My wish list:
(1) More than 20 Y values per X value
(2) Better text editing -- direct in the plot and not in a text window;
more flexibility in choice of styles, fonts, super/subscripts, etc.
(3) Plot styles/page layouts should be saved with the plots (e.g., if you
save a plot with Page Setup in landscape, it should stay that way next
time you open the plot)
(4) Somewhat minor--on our PM7500, KG doesn't seem to be aware of changes
to the monitor resolution; you have to quit and re-launch KG to be able to
use the expanded screen area.
(5) Easier macro/calculator languages, or better manuals for the existing ones
(6) Make the link between the column titles and plot axis labels
dynamic--currently, if you change a column label, you have to make a new
plot to get the changes to take effect
(7) Drag-and-drop awareness--e.g., drag a column from the data window
onto a plot to modify the plot--drop the column on the Y axis to add a set
of Y values, drop it on the X axis to change the X values. Feasible? I
have no idea, but I would find it a useful feature.
(8) Allow axis breaks.
Cheers,
Raghu <ra...@u.washington.edu>
Michael B. Moore
In Article <ej-230996...@d251-powmac9500.stanford.edu>,
e...@white.stanford.edu (EJ Chichilnisky) wrote:
>6. Allow copying of columns of data with titles (that is, if you copy
>several columns from one data sheet into another, the titles should come
>with the data!).
>
Hold down the option key while performing the copy to clipboard. The column
titles will then be copied with the data.
>EJ Chichilnisky
>e...@white.stanford.edu
r...@po.cwru.edu
In article <527ast$2...@ucsbuxa.ucsb.edu>,
Robert Herrick <650...@ucsbuxa.ucsb.edu> wrote:
>In <skok-23099...@news.uni-stuttgart.de> sk...@itw.uni-stuttgart.de (Holger Skok) writes:
>
>>In article <rjbailey-200...@rjbailey.sc.scruznet.com>,
>>rjba...@scruznet.com (Jeff Bailey) wrote:
>
>>>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
>>><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>>>Mac is in the works, and is inviting user suggestions for what we'd like
>>>to see in the next version. I've already sent him mine, and he would like
>>>to see suggestions from other users as well. Development for the Mac has
>>>been stalled while the Windoze version of KGraph was cranked out, but now
>>>it's time for a new Mac version.
>
Add:
The ability to color points according to values from a selected
column.
Rob Lake
Environmental Modeling Inc.
r...@po.cwru.edu
Jim Jennings
>
>Add:
>
>The ability to color points according to values from a selected
>column.
>
Ditto. Right now the only Mac program I know of that does this is
Spyglass Plot (now Fortner Plot I suppose). However, Spyglass Plot does
not measure up to Kaleidagraph in most other aspects, IMHO.
.........................................................
Jim Jennings
Research Associate jenn...@mail.utexas.edu
Bureau of Economic Geology (512) 471-4364 (voice)
University of Texas at Austin (512) 471-0140 (fax)
.........................................................
Jon Guyer
In article <jenningsj-250...@begmac33.beg.utexas.edu>,
jenn...@mail.utexas.edu (Jim Jennings) wrote:
> Ditto. Right now the only Mac program I know of that does this is
> Spyglass Plot (now Fortner Plot I suppose). However, Spyglass Plot does
> not measure up to Kaleidagraph in most other aspects, IMHO.
Igor Pro can assign color, symbol size, and symbol based on three distinct
columns of data. Very handy.
--
Jonathan E. Guyer
Join LMESOTENPAMSNFZ
("Let's make everything south of the 89th parallel
a Microsoft no-fly-zone" AKA "Give Bill the Pole!")
Dave Jaksha
It would be nice, not, to have the 32K limit in the data window. I run
into this all the time, and have to move to IGOR to handle the large data
sets.
Dave
Steve Byan
I vote first for better Applescript support. I'd particularly like to
script Kaleidagraph to produce a series of plots which I would then
combine into a QuickTime visualization.
I also have need of half-plane polar plots; I'd like to see that type
added, or at least some way to add a clipping region to a plot so that I
could manually edit a full-plane polar plot down to a half-plane plot.
Regards,
-Steve
--
Steve Byan internet: st...@hi.com
Hitachi Computer Products (America), Inc.
1601 Trapelo Road phone: (617) 890-0444
Waltham, MA 02154 FAX: (617) 890-4998
Kevin R. Boyce
>In article <52a6em$r...@madeline.INS.CWRU.Edu>, r...@po.cwru.edu wrote:
>
>>
>>Add:
>>
>>The ability to color points according to values from a selected
>>column.
>>
>
>Ditto. Right now the only Mac program I know of that does this is
>Spyglass Plot (now Fortner Plot I suppose). However, Spyglass Plot does
>not measure up to Kaleidagraph in most other aspects, IMHO.
Igor Pro does that too. It's necessary for what I've been doing lately.
--
keV Kevin....@gsfc.nasa.gov
We work hard.
We play hard.
And you wouldn't *believe* what we do to microwave ovens.
Yongsup Park
Please make the nonlinear curve fitting routine more powerful and flexible.
Windows programs like Origin does it much better than KG.
For example, temporary parameter freeze, settable range, and more than 9
parameters. Get the latest version of Origin and study it !
Or you can take a look at IgorPro.
Due to lack of power and flexibility in curve fitting, now I am seriously
considering switching to IgorPro or Origin on my Pentium box even though I
have been using KG for almost 7 years now.
--
************** Note that my office number has been changed ************
Yongsup Park (박 용 섭) | 716-275-[7738(O) 6987(L) 8527(F)]
Dept. of Physics and Astronomy | pa...@alfalfa.pas.rochester.edu
University of Rochester | http://www.pas.rochester.edu/~park/
Jon Guyer
In article <52cpa5$2...@en.com>, bri...@en.com (James Bridges) wrote:
> I agree this would be nice in the Open dialog box, but I think this is
> more a limitation of the Apple Toolbox routine. If this is really a
> problem, just get into the Finder, shift-click all the files you want
> to open and double click on one of them. KG will open them all.
Or use 7.5's find tool. Integration of this thing into the OS, a la BeOS,
would be really cool. The SFGetBox is decent... by 1984 standards, but I'd
say it's due for a major overhaul... and I don't mean Now's and Norton's
kludges.
Scott A. Schroeder
In Article <rjbailey-200...@rjbailey.sc.scruznet.com>,
rjba...@scruznet.com (Jeff Bailey) wrote:
>Bill Jones <mailto:jo...@synergy.com> at Synergy Software
><http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>Mac is in the works, and is inviting user suggestions for what we'd like
>to see in the next version. I've already sent him mine, and he would like
>to see suggestions from other users as well. Development for the Mac has
>been stalled while the Windoze version of KGraph was cranked out, but now
>it's time for a new Mac version
1) 3D plots (contour, density, surface) with Mathematica style flexibility
2) Variable column widths and row heights within a data file
3) More flexible text editing (mathematical symbols without using an alt.
font)
__________________________________________________________
Scott A. Schroeder \ "Views are my own, Standard Disclaimer, etc."
Rockwell Science Center \ Scott_S...@scimail.risc.rockwell.com
Thousand Oaks, CA 91360 \ http://www.risc.rockwell.com/
Kelly Miller
And another, which I remembered today, while looking over some old data:
Allow more than one data file to be opened in the "Open..." dialog box
(through multiple selections). It's a pain when, like today, you want to
open 3 or 4 data files, all in the same folder (and buried half a dozen
deep in the catalog heirarchy), and have to repeat the Open command for
each file.
Kelly Miller
mil...@aries.scs.uiuc.edu
James Bridges
In article <52cnc9$6...@vixen.cso.uiuc.edu>
mil...@aries.scs.uiuc.edu (Kelly Miller) writes:
> Allow more than one data file to be opened in the "Open..." dialog box
> (through multiple selections). It's a pain when, like today, you want to
> open 3 or 4 data files, all in the same folder (and buried half a dozen
> deep in the catalog heirarchy), and have to repeat the Open command for
> each file.
I agree this would be nice in the Open dialog box, but I think this is
more a limitation of the Apple Toolbox routine. If this is really a
problem, just get into the Finder, shift-click all the files you want
to open and double click on one of them. KG will open them all.
James Bridges
Ray Osborn
Yongsup Park wrote:
>
> Please make the nonlinear curve fitting routine more powerful and flexible.
> Windows programs like Origin does it much better than KG.
> For example, temporary parameter freeze, settable range, and more than 9
> parameters. Get the latest version of Origin and study it !
> Or you can take a look at IgorPro.
>
> Due to lack of power and flexibility in curve fitting, now I am seriously
> considering switching to IgorPro or Origin on my Pentium box even though I
> have been using KG for almost 7 years now.
>
<http://www.cherwell.com/cherwell/ProdHome/proFit.html>. A demo has
been available for a few months now although I think that it will expire
at the end of September. In the short time that I've been using it,
I've been extremely impressed by the way they have thought things
through. They have avoided, it seems to me, most of the idiosyncracies
which have made Kaleidagraph awkward to use, while providing all the
functionality. For example, I never used Kaleidagraph for much more
than producing publication plots because of the arcane method of writing
complex functions. On the other hand, proFit uses a Pascal-like
language with enough built-in procedures to make it easy to customize
the working environment. The fitting routines are an integral part of
the package with a preview window that makes setting up initial
parameters particularly simple. It's also Apple scriptable.
For the record, I'm in no way connected to Cherwell Scientific.
--
------------------------------------------------------
Dr Ray Osborn Tel: +1 (630) 252-9011
Materials Science Division Fax: +1 (630) 252-7777
Argonne National Laboratory E-mail: ROs...@anl.gov
Argonne, IL 60439-4845
[Please note change in telephone area code]
James T. Platt
I recently checked out the Kaleidagraph demo because I had not
been satisfied with the way CA-Cricket Graph and StatView handles a
kind of graph commonly used to present chemotherapy data.
Unfortunately, the Kaleidagraph demo I tried does not seem to be any
better.
What I want to do is overlay line graphs with error bars and
measured scales on both axes. This seems simple but it is actually very
difficult in every graphing program I have tried. Let me give you an
example:
X-axis - days since chemotherapy began.
Y-axis - Mean tumor size in cubic mm +- standard error.
The different lines on the graph are for different treatments and
control groups. This is a standard method for presenting chemotherapy
data.
CA-Cricket Graph (CG) will do all this as long as you compute the
mean and standard error with another program (such as StatView) and
transfer the data to Cricket Graph. You then need to assign the data
columns containing the errors with the correct columns containing means
and the rest is easy. CG can compute statistics itself but not in a
form that is convenient for plotting error bars. CG will do the job
but it takes a lot of work to get there.
StatView (SV) will conveniently take the raw data (entered in one
column for each day for each experimental group), automatically compute
the mean and standard error and plot a line graph with the means as the
y-values for points and the standard errors for the error bars but it
will not measure out days on the x-axis. It puts the column names,
evenly spaced, on the x-axis - there is no way to give x-values to
points that have error bars. In other words, it puts just as much
space between the points for day 18 and day 20 as between the points
for day 20 and day 25.
The SV method for overlaying graphs is to just make separate plots
and pile them on top of each other other in its graphic editing
interface (the view window). Overlaying graphs this way is
unacceptable to me because lining up the graphs perfectly requires
looking at the plot at a high magnification, such as 800%, and printing
to check that things are lined up just so. If any changes have to be
made after that, the whole process has to be repeated.
The Kaleidagraph method has similarities to both CG and SV but
could use more features. All that is needed is a way to assign an
x-value to each column of y-data and automatic calculation of column
statistics for plotting.
James Platt
Yale University
Chris Cooper
Re Kaleaidagraph,
I hate to get all user friendly on you guys, but the biggest problem in
Kaleidagraph is its non-intuitive interface. It may as well be a DOS program.
It's the only current Mac program that I can think of still shows you control
codes on the screen for subscript and superscript for example.
More importantly why use the formula
bar and have to remember column numbers to do simple spreadsheet-like functions
on your data? Surely it is easier to click on the column and type an equation?
If I want to let my students and Post-Docs loose on this program
they have to read the manual quite closely. This is clearly not the case with
Cricket Graph for example. Also the scripting required for curve fitting is
much more complicated than necessary. Deltagraph, Sigma Plot and Multifit/
Ultrafit handle this much better (and have better statisitics to boot).
Also as mentioned previously the output of graphs on a page for printing is
very restrictive.
So why do I use Kaleidagraph? Well it is clearly much faster than Excel and
Deltragraph etc. and therfore much better with large data sets. It is smaller,
requires less memory and can run on my slow and fast Macs. But I would buy 3
more copies for my lab, rather than the one I use, if it was easier to use.
To sum up if I was revising Kaleidagraph I would:
By all means keep the scripting options which I presume the more technical
people out there like but:
Look at the basic interface from Cricket Graph (1.3 was the best, but even II is
OK) for column manipulation, dialog boxes for plotting and presentation of
output.
Add more basic statistics functions (like those in Excel, which must be very
easy to code for, and you could hardly do it slower than Excel's analsis
toolpak).
Look at the non-linear fitting from Multifit or Ultrafit packages in terms of
how they are presented. A simpe dialog box should come up with an editable
equation that can be saved and re-used time and again easily.
Chris Cooper
University of Essex
Michael B. Moore
In Article <paul.savage-ya02318...@newshost.its.csiro.au>,
paul....@chem.csiro.au (Paul Savage) wrote:
>I find KaleidaGraph to be one of the most difficult
>Mac programs to work out (that and VersaTerm Pro). Also a single,
>well-written, WELL INDEXED manual would be nice.
>
>I'll give you an example. I wanted to figure out how to use date formats
>for my X axis so I looked up "date formats" in the index to the
>KaleidaGraph Learning Guide (3rd edition). It says page 136. On page 136 is
>"Macros" and nothing about dates at all. In the index is "date formats -
>controlling axis label display" on page 198 -- sounds good. On page 198 is
>"Using Subscripts, Superscripts, and an Alternate Font". Nothing about
>dates. What the hell is with this manual???
The index lists page numbers for both the reference guide and the learning
guide. One of them is always shown in italics; don't remember which.
>
>Paul Savage.
Ummm, I've seen manuals that are *much* worse than Synergy's. Have you ever
tried to read a Microsoft manual? Complete nightmare. I'll agree, though,
that the manuals aren't as clear as they ought to be. Fortunately, Synergy
has one of the best tech support lines I've used.
As for KG being hard to work with, I guess I don't agree, overall. It *does*
have some unusual idiosyncrasies, like how it handles data and plot files
when you want to modify data, but it's pretty powerful, and fast, and is
relatively inexpensive.
BTW, do any of you use VersaTerm Link for email and newsgroups? I really
like it. In fact, I would buy VT-Pro just to get VT-Link. Easy to configure,
easy to use. I guess I'm a satisfied customer of Synergy's.
Mike Moore
NASA/MSFC
Michael B. Moore
In Article <52gvj0$6...@news.ycc.yale.edu>, pl...@biomed.med.yale.edu (James
T. Platt) wrote:
> What I want to do is overlay line graphs with error bars and
>measured scales on both axes. This seems simple but it is actually very
>difficult in every graphing program I have tried. Let me give you an
>example:
>
> X-axis - days since chemotherapy began.
>
> Y-axis - Mean tumor size in cubic mm +- standard error.
>
> The different lines on the graph are for different treatments and
>control groups. This is a standard method for presenting chemotherapy
>data.
>
> The Kaleidagraph method has similarities to both CG and SV but
>could use more features. All that is needed is a way to assign an
>x-value to each column of y-data and automatic calculation of column
>statistics for plotting.
>
>James Platt
>Yale University
James, I'm not sure I'm reading you completely, but I *think* you can do
this in KG. KG has the ability to plot 9 different XY pairs, i.e., you don't
have to use the same x-column for all of your y's. As for the standard
error, KG *will* do this; you just need some documentation to show you how
to do it.
Next time you try the demo, do this:
1. Input your data in the data window,
2. Choose Line plot from the Gallery menu,
3. In the resulting dialog window, notice that there's a square drawn
around the X. Click and hold on that square; a list running from X1 to X9
should result.
Hope I've helped a little.
Mike Moore
NASA/MSFC
Sergio Calatroni
New Required features for KaleidGraph?
Those that first come to mind are:
1- 3-D Plots
2- Capability of defining columns as automatic calculations on other
columns: I mean, not like the plain formulas, that calculates and
fills the target column with the results, but really a dynamic method
that recalculates if the source data are changed (Excel-like I would
say)
3- A more straightforward implementation of the plot sizing routines
4- Expand the fitting routines
5- FFT will be useful (Origin docet)
Sergio
Erik Fabry
In article <rjbailey-200...@rjbailey.sc.scruznet.com>,
rjba...@scruznet.com (Jeff Bailey) wrote:
> Bill Jones <mailto:jo...@synergy.com> at Synergy Software
> <http://www.synergy.com> tells me that a revision of KaleidaGraph for the
> Mac is in the works, and is inviting user suggestions for what we'd like
> to see in the next version.
Suggestions for new KG version:
1 Add EPS export
2 Possibility to save printing plot size and page layout
_____________________________________________________________________
Erik Fabry fa...@vki.ac.be
Von Karman Institute for Fluid Dynamics http://www.vki.ac.be/
_____________________________________________________________________
Bjorn Nicander
> ... possible improvements to KaleidaGraph.
The interface would win by becoming more Mac-like.
1. Direct editing of any text in the plot window so the Text tool dialog
box can be removed from the program.
2. Show line style, marker etc for each column in a new row beneath the
title row in the data window. To change, double click the cell to get to
the Plot Style dialog.
3. A few slide and overhead layout features would be nice:
- a feature to make a window have the right proportions
- some kind of automatic center in window function for text and graph
- checkboxes to turn on superimposed boxes with the right proportions in
Show Page
Dr. Bjorn Nicander
Department of Plant Physiology
Swedish University of Agricultural Sciences
P.O. Box 7047, S-750 07 Uppsala
phone +4618671493 fax +4618672930
Jon Guyer
In article <fabry-01109...@macfe.vki.ac.be>, fa...@vki.ac.be (Erik
Fabry) wrote:
> 1 Add EPS export
LaserWriter 8?
Jeff Bailey
FFTs are already available as KG macros. There's one on the Info-Mac
archive, or I can send you a copy if you wish. (It flies!)
-Jeff
In article <325101...@mail.cern.ch>, Sergio Calatroni
<sc...@mail.cern.ch> wrote:
--
Jorg Helmut Kleinschmidt
I would like to see a possibility to save all created data and plots
including style and page layout information into one file. This makes
it much easier to print the same thing at a later time once again.
Currently, one has to remember all the file names, again put the plots
together on the layout page to eventually get the same printout as
before.
Craig Jones
In Article <skok-02109...@news.uni-stuttgart.de>,
sk...@itw.uni-stuttgart.de (Holger Skok) wrote:
>In article <fabry-01109...@macfe.vki.ac.be>, fa...@vki.ac.be (Erik
>Fabry) wrote:
>
>It doesn't seem to be possible to export them other than as PICT. So if
>I want to get at the slope of a linear fit, for example, to use it in further
>processing, I have to manually pluck the PICT apart, select the desired
>item of text, copy and paste and finally get rid of that pesky "x" which
>sits directly to the right of the numerical value that I am intested in.
>That way I always select the x, too, when double clicking on the piece of
>text.
>
>It would be extremely nice, to be able to export the results of any of
>the fits as tab-delimited data or at least to have the option of
>putting them into a new data window.
>
This is already there. Results of the last curve fit are in the memory
variables m1, m2, ... and can be used in the Formula window right away.
Also, if you reselect the fit from the curve fit window, you will see a
checkbox for "view"--checking that will bring up a little window with the
parameters from that fit and buttons to copy to the calculator or the clipboard.
Of course, this can get tiresome if you do this for many fits (some batch
access would be nice).
Craig Jones cjo...@mantle.colorado.edu
Research Associate, CIRES, University of Colorado, Boulder
WWW: http://cires.colorado.edu/people/jones.craig/CHJ_home.html
Craig Jones
In Article <DynqD...@murdoch.acc.Virginia.EDU>,
the current setup: add a button to the layout page (which really shouldn't
be a modal dialog) for "Save as Graph File..."; when you'd reopen that file,
you get all the plots (with the data hidden, as at present for graph files
with data) and the page layout would also appear (and maybe even have a
document name!).
Boyan Boyanov
In article <skok-02109...@news.uni-stuttgart.de>,
sk...@itw.uni-stuttgart.de (Holger Skok) wrote:
>In article <fabry-01109...@macfe.vki.ac.be>, fa...@vki.ac.be (Erik
>Fabry) wrote:
>
>>In article <rjbailey-200...@rjbailey.sc.scruznet.com>,
>>rjba...@scruznet.com (Jeff Bailey) wrote:
>>
>>> Bill Jones <mailto:jo...@synergy.com> at Synergy Software
>>> <http://www.synergy.com> tells me that a revision of KaleidaGraph for the
>>> Mac is in the works, and is inviting user suggestions for what we'd like
>>> to see in the next version.
>
>Another request concerns the way of handling curve fitting results.
>It doesn't seem to be possible to export them other than as PICT. So if
>I want to get at the slope of a linear fit, for example, to use it in further
>processing, I have to manually pluck the PICT apart, select the desired
>item of text, copy and paste and finally get rid of that pesky "x" which
>sits directly to the right of the numerical value that I am intested in.
>That way I always select the x, too, when double clicking on the piece of
>text.
>
>It would be extremely nice, to be able to export the results of any of
>the fits as tab-delimited data or at least to have the option of
>putting them into a new data window.
I think you can do this now with the formula window. For the linear fit
that you explicitly mention:
(a) select the data that you want to fit to
(b) fit a line to it
(c) assuming that the data is in column 1, the slope of the line is
lin(1,c1)-lin(0,c1)
The result will be displayed in a dialog box on the screen and you
can copy it to the clipboard if you so desire. You can play similar
tricks with any othe fit in KG.
I admit that this is rather clunky, but so is the rest of the KG
interface :-). However, those of you who think KG's interface is
the worst of the Mac breed of programs should check out Igor. Now
this is clunky (though powerful)!!!
Cheers,
Boyan
-----------------------------------------------------
Boyan I. Boyanov | phone: (919) 515-1330
NCSU, Physics, Box 7518 | fax: (919) 515-4496
Raleigh, NC 27606-7518 | email: boy...@eos.ncsu.edu
-----------------------------------------------------
<http://ncstarg.physics.ncsu.edu/boyan>
-----------------------------------------------------
Boyan Boyanov
Here is my wish list for KG:
(1) First and foremost I need some kind of a plugin architechture so I can
write my own data processing modules in C and use KG to visualize the
results. The current RPN thingy is for people whose heads are even more
pointed than mine :-).
(2) Something should be done about the ***horrible, horrible, horrible***
way in which KG currently handles roundoff error. Here is an example (KG
3.0.5/PMac):
(a) create the following series in c0:
0.000000000
0.2500000000
0.5000000000
0.7500000000
1.000000000
(b) in the formula window type
c1=c0*5.4310
(c) the results in c1 are
0.000000000
1.357750058
2.715500116
4.073249817
5.431000233
As you can see, KG does not seem to be able to multiply 1*5.4310! Note
that setting the column data types to Double has no effect on the result.
Does
anybody know of a workaround?
Regards,
Stuart Craig
>Another request concerns the way of handling curve fitting results.
>It doesn't seem to be possible to export them other than as PICT. So if
Goodness-grief! Read the manual. With the plot displayed, select the
curve fit you're using (under the Curve Fit menu). Click on the View
check-box for the curve you want, then click on the Clipboard button.
You now have a copy of the curve fit parameters in the clipboard.
However ... I'd like to see the original cubic-spline macro put back onto
the distribution disks since this macro produces columns of the spline
parameters. // Stuart Craig
Robert Hohlfelder
In article <boyan_boyanov-ya023080000210960911010001@news>,
boyan_...@ncsu.edu (Boyan Boyanov) wrote:
> Here is my wish list for KG:
>
> (1) First and foremost I need some kind of a plugin architechture so I can
> write my own data processing modules in C and use KG to visualize the
> results. The current RPN thingy is for people whose heads are even more
> pointed than mine :-).
Ditto!
A revised macro language would be great. The "RPN thingy" is painfully
obscure; I played with it one day for about fifteen minutes and then
abandoned it.
- Robert
--
Robert Hohlfelder
r...@leland.stanford.edu http://www-leland.stanford.edu/~rjh/
(415) 723-1284
Dave Jaksha
In article <rjh-031096...@nixdonhomer.stanford.edu>,
r...@leland.stanford.edu (Robert Hohlfelder) wrote:
Ditto again. I would love to build scripts, but the learning curve is
just not worth it.
Dave
Scott E. Lasley
In article <skok-07109...@news.uni-stuttgart.de>, sk...@itw.uni-
stuttgart.de (Holger Skok) writes:
> I'd like to be able to just grab the displayed equations and copy them
> to another application as TEXT or - much nicer - have an
> Applescript command to just extract the fit parameters as TEXT, or,
> export them into a new data sheet.
if you select the text tool from the tool palette and double click on the
displayed equation, KG will bring up the standard text edit dialog box
containing the equation. you can edit your equations there and copy them to
the clipboard as text. i'm not sure if this process is scriptable or any
easier than the process you currently use.
hope this helps,
Scott E. Lasley sla...@space.umd.edu | http://space.umd.edu/
"Push the button, Frank." MST3K | http://umtof.umd.edu/pm/
Lawrence D'Oliveiro
Here's a general suggestion for all developers of specialized graphical
applications (not just Kalediagraph): allow export of your graphs as
QuickDraw GX <http://www.ixmedia.com/quickgx/> graphics.
This gives a high-quality *fully-editable* format that can be imported
into applications like Lightning Draw
<http://www.larisoftware.com/Products/LightningDraw.html>, tweaked a
little bit (or a whole lot :-)), pasted in fully WYSIWYG form into word
processors and the like, and printed at full resolution on both PostScript
and non-PostScript printers.
John Williams
Chris Cooper wrote:
>
> Re Kaleaidagraph,
>
> I hate to get all user friendly on you guys, but the biggest problem in
> Kaleidagraph is its non-intuitive interface. It may as well be a DOS program.
> It's the only current Mac program that I can think of still shows you control
> codes on the screen for subscript and superscript for example.
>
parts deleted
>
> Look at the basic interface from Cricket Graph (1.3 was the best, but even II is
> OK) for column manipulation, dialog boxes for plotting and presentation of
> output.
parts deleted
>
> Look at the non-linear fitting from Multifit or Ultrafit packages in terms of
> how they are presented. A simpe dialog box should come up with an editable
> equation that can be saved and re-used time and again easily.
I agree with this. We need a package to replace Cricket Graph (1.5.1)
which will have a simple, easy-to-learn interface. Personally, I like KG
but the interface badly needs overhauling to render it easier on the eye
and simpler to learn. Could we have an option for shart or long menus
which would flip from a basic graphing tool to an advanced one?
A personal request is to include the 4-parameter logistic fitter for
sigmoid dose-response curves. I managed to program this in the curve
fitting menu but it's not easy to use. Each new data set requires you to
go through a cascade of dialog boxes to find the precise places where
the initial parameter values are stored, one example of how the
interface can be awkward to use.