Recommend a Graphing Program?

[Groth]

Do you have a favorite way of turning a matrix of values into a pretty picture?

I need a better way of making x-y plots with color-coded values. If it can do the same sort of display on the surface of a 3D model, so much the better.

What works? What's easy to use? What's cheap or free?

[BillA]

PM at you
easy to use ???

TechPlot
Surfer by Golden

difficult to believe you have exhausted the forms of Excel

[Groth]

Don't have excel. Didn't come with the last 'complete system' I bought, and I ain't planning to give Macroslut any cash.

[BillA]

well, Excel is far easier than those others,
far easier

what OS ?

[Groth]

I've Debian Linux, Win98, and WinXP. I'll continue to use the MS products I have, I just don't plan to buy more.

[pHaestus]

microcal origin

isnt gnuplot tolerable for *nix?

[HighFlowRod]

Curve Expert

[Groth]

I tried gnuplot a couple years back, found it unwieldy. I'll check it out again, plus the others mentioned. My brain aches already.

In other, unrelated news (hijack my own thread?), it took me all of 15 minutes to make a corner tapped orifice flowmeter using two pieces of copper pipe, an old penny, a straight pipe coupler, and a bit of small diameter brass tubing.

Now to see if it's of any use....

[pHaestus]

Origin's spendy ($500US I think for academics) but very good and intuitive. Elitist chemists use it or so I've heard.

[Groth]

Elitist chemists? LOL, implies there's another kind.

Yeah, Origin looks to be waaaay out of my price range. Some nice lookin' contour color-fill graphs. Damn, the whole thing looks pretty -- I'll have to play with the demo even though I can't buy.

[Incoherent]

I'll lend my support to Excel. Understand the reluctance to pay MS but it's more capable than most people, esp. Phds, realise.
Matlab is very good too, but not cheap or easy.

The flowmeter. Groth, beautiful,I love it. I'm making one.

Incoherent

[Groth]

'Twas inspired by Bill and his bfh. The majority of the time was spent finding an old enough penny (can't be using one of the modern zinc buggers).

Did a bit playing with gnuplot. I can get contours, but not the pretty shaded color fills I want. I'll give it a little more effort, then move to the next.

In the meantime, a 2d coldplate. 40 by 6 mm, 10 mm wide uniform heat current in, 311.7 K peak temp, arbitrary convection coefficient leading to a 300K isothermal 'great beyond'. Check out them beautious isotherms!

Very, very (very!) rough start.

[bigben2k]

Can you detail the scale here?

[Groth]

What kind of details are you looking for Ben?

[bigben2k]

Well, you wrote "40 by 6mm" and I see 8 by 5 sections in the graph. Also, you mention 300K to 311.7 K spread, but I only see 7 lines (forming 8 sections) for the whole 11.7 deg spread.

[Groth]

The 6 mm thickness shows up as two edges + four grid lines. In a similarly odd way, the 40 mm width is 8 major grid lines * 5 ticks each. It includes the left edge, but not the right. :shrug: Confusing artifacts of the graphing program -- I should have turned the grid off, but I was using 'em to make sure the aspect ratio was correct.

The isotherms are integers. There's only seven because the cold, far corners are at 304.5 with the h I was using. When I tried to get it to label the contours, the labels overlapped each other and the graph.

Anyway, that model is a day old and already obsolete.

Here a different set-up, with two sets of isotherms corresponding to two different convection coefficients. I'm curious to whether the differing shapes are real or model inaccuracy.

[Les]

[quote=Groth]I'm curious to whether the differing shapes are real or model inaccuracy [quote]

Which/who's model?

[Groth]

My model (in progress). Leverages the readily available SPICE simulators; voltages for temperatures, currents for heat flow, capacitance for specific heat, resistors for thermal impedances, etc. etc. Half the fun is figuring how to build and visualize netlists with thousands of nodes.

Of course, valid answers would be cool, too. I'm playing with different wiring configurations to see what might work. No results yet, just purty pictures.

[bigben2k]

This might be up your alley Groth: you might be able to find something within one of these sections: http://www.openscience.org/index.php?section=23

[Zogthetroll]

try looking around for Graphical Analysis for Windows, I used that a while back, but i'm not sure where to get it at the moment. its not bad.

[Groth]

So many programs, so little brainpower to use them with...

The demo of Origin has finally begun to cooperate. Here is a 3D color contour plot of the temperature of a square plate with isothermal edges and a central heatsource.

The peak is too sharp -- a combination of the model's granularity and continuing suckage.

[Groth]

Discovered something cool today. SPICE simulators can make .wav files (so you can listen to simulated audio circuit) and Origin can import them. Origin doesn't quite know know to handle a 1280 channel .wav, but with a little scripting to chop it up it becomes useable.

I'm unskilled with Origin and have never made an animated .gif before, so quality is low...Here is a proof of concept movie for transient modeling. [Edit: gif removed]

A 40 by 8 mm plate; a 10 mm 10 watt source; and a 1 watt 0.1 second off center heat pulse.

Edit: I made a mistake on the thermal conductivity, so the temperature numbers are wrong wrong wrong. But the un-numbered version is still neat and one post down.

[Groth]

I chopped the edges off the animation. It lacks a scale now, but is much easier to watch. http://pages.sbcglobal.net/water.gro...es/pulse_b.gif