I apologize for going to miss the class tonight. I was so looking forward to this review.

If if it really long, You can also set it to capture 1 frame every couple of seconds for a time lapse effect. It'll make the ecotect analysis demo seem much quicker in the video.

Simon

Thanks!! :)

I fixed it for you. I embedded the video code into your page. Check it out!

Best
-Mohamed

Anyway, if it is possible, please review my grasshopper file and let me know if I made any mistake or missed some problems.

Thanks! and See you tomorrow :)

Was it just too slow? Maybe you can post an iteration with the solar analysis being computed?

Nate

It takes so long time to finish running Galapagos and Ecotect and I captured some process. I'm not sure if my definitions are all correct.

Please review and let me know if my definitions are incorrect and I missed some requirements.

I'm so sorry that I have to miss the last class.
I learned a lot about Grasshopper and Galapagos and I really enjoyed the class.

Thank you very much.

Please follow the format described here.

Some notes:

1. You must use Vimeo or YouTube to post the video. Please do not upload video files to this wiki as space is limited. Going through these sites will also allow you to embed the video file directly to the wiki page.
2. Final projects must be submitted on time to this wiki (no later than Thursday 6:00 PM). It will be considered late if it is after that point..
3. Uless you have already received my pre-approval, all students are expected to attend the discussion review at NBBJ. This is a key part of your participation.
4. I will be selecting certain projects for discussion tomorrow. I will contact you in advance if I have selected your project. Be prepared to share describe your project to the group.

Please post any general questions about the final submission here (Format, schedule, etc)….

I don't know how compatible these two systems are directly… basically the system will need to look something like:

1. A series of points control the Voronoi cells.
2. Each point will have a square area associated with it (the target fitness).
3. You will give Galapagos control over the X and Y position of each individual point The genes).
4. Galapagos will move the points around until the target fitness is achieved.

This, of course, is probably an over simplification of the approach. Give it a shot and post your attempts.

Enjoy
-Mo

We did this in class for the other components. you will have to do it again for this component.

but.. I have problems when I run the program…

I tired to apply the definition you created to my file, but after that, there is nothing coming out on Ecotect screen. (There is no data coming out) I checked the definitions but couldn't find what was the error. I am not sure if my connecting definition is wrong or other issues exist.

Therefore, for now, there is only two objectives affect the result. How can I add this Ecotect result as my third objective?

Could you check it for me? Thanks…

[[html]]
http://theprovingground.wikidot.com/local--files/won-hee-ko/0323.3dm,
http://theprovingground.wikidot.com/local--files/won-hee-ko/0330.ghx
[[/html]]

Let me know if this resolves your problem!

Also looking at the "program.ghx" file i am not sure how to change the desired area quantity there are two panels one on the very far left that can be edited and the other on the far right that states "Desired Area" that can't be edited. I understand how both these definitions work I need help bringing them together. Any help would be awesome. Thanks.

The components return lists of True or False boolean values which you can then use to "Cull" your list of panels and remove those not >x. Plug your panels and your conditional result into the "Cull Pattern" component to achieve this.

The definition is here: Triangle Panels for Ecotect

The definition is here: Triangle Panels for Ecotect

It's weird that TriGrid doesn't extend itself to the boundaries of the rectangular surface, perhaps this is because it will not/cannot create a half equilateral triangle - only whole equilateral triangles. I know that might be stating the obvious, but hey…

In any case, I'll search the Grasshopper forums later, I'd be surprised if someone hasn't encountered this issue in the past.

I'll work on adding color, but that's less of a priority as filling the rectangle right now, imo.

As far as the avi goes, I'm assuming that you opened the video with quicktime or windows media player? It didn't look super crisp, but was still clear/totally legible. This one was really just a test to make sure I could produce one.

Lastly (and this is sort of nitpicking), I've played around with the distance of the "points of influence" (Hot Spot and Best View) and I can't seem to achieve greater sensitivity with regard to the TriGrid Subdivisions.
In other words, each "core" triangle will subdivide, but when you move the tolerance slider in increments of one integer to achieve this, oftentimes two or three of the core TriGrid triangles will subdivide simultaneously (as opposed to one at a time/one by one). Any thoughts on how to get these core triangles to subdivide one at a time? I'm thinking it has to do with the formula used to measure the tolerance and will explore that in more depth, if I'm wrong about this give me a hint..

Looks like you are on the right track… although the video file playback is behaving strangely on my computer. From what I can see, it looks like it is working

You can incorporate color by using the custom preview component (under Params) and the Colour components (under vector)

The perimeter triangles using the method of subdivision may be tricky I will have to investigate how to accomplish this.

So, I ended up setting the Fitness value to 5.5, which would mean that we are equally weighting solar shading vs. view…maybe this can be something more spicy?

Here is the "First Run" document set as a zip file (avi video, pics of start and final solution, and grasshopper file), I would have embedded this but I don't want to fill up my Vimeo account just yet.

Please have a look and give me suggestions to make this more interesting.

Here are my thoughts on how to do so…

1. Make 2 SubD TriGrid Rectangle surfaces of different sizes so it looks like a building corner.
2. Change up the Fitness, give more priority to View and then to shading.
3. Could we incorporate color? Black, White, Grey? Maybe I will go Photoshopping…

Also, someone should make a disclaimer about how addictive G&G can become. I mean, I'm swamped with projects and watching the solver run is a little hypnotic…

[[html]]
[[http://theprovingground.wikidot.com/local--files/suhee-jung/110329_MaximumVolumn%281%29.3dm]]
[[http://theprovingground.wikidot.com/local--files/suhee-jung/110329_panel.ghx ]]
[[/html]]

Thank you

-Nate

-nate

It would take long time to run Ecotect and I want to confirm my all definitions are correct before running Ecotect.

Please let me know.

[http://theprovingground.wikidot.com/local--files/myoboon-hur/ARCH517_FinalProject_Myoboon_20110329.ghx]

Thank you.

2) For Ecotect, I will try to set up independent controls for each face and if I have a trouble, I will ask you again.

Thank you.

I also tried to apply the ecotect definition to my grasshopper file, butit has same issue with Myboon-weird geometry on Ecotect Screen.. is it ok?…

I wanted to have three different objectives for the optimization of my building; Minimize total facade area, Maximize south facade area, Minimize solar radiance via ecotect analysis. When I run this, it took time forever… I guess it is because of too much slides and definitions are going on this file. How can I develop or refine this more?

[[http://theprovingground.wikidot.com/local--files/won-hee-ko/0323.3dm,
http://theprovingground.wikidot.com/local--files/won-hee-ko/0328_ecotect5.ghx]]

Please take a look and give me some advise! :)
Insert any HTML code, including widgets and video or audio players
[[/html]]

1. I've got two "Attractor Systems" which I've renamed "Hot Spot" and "Best View". They are positioned and the coordinate on the circle should remain constant.

2. The Slider controlling "Tolerance" when increased will create a denser TriGrid SubDivision towards the "Hot Spot" point. I've done the opposite with the "Best View" point/"Tolerance" slider - in other words, increasing it's value results in a less dense grid.

3. I'm assuming that these two sliders are my "Genes" but I can't seem to connect both sliders to the Galapagos "Genome". Also, I have no idea what my "Fitness" should be yet…

Help…

Here is the Grasshopper Definition

Here is a screen capture describing the sliders (open it in a new window to see the whole thing).

Keep in mind everything pretty much works how I hope it would, I just need some tips on how to play the two sliders against one another…

There are several problems that I see with the definition…

1. Genes: A single slider is used to control all surfaces making multi-objective optimization impossible. You should have multiple sliders controlling the different faces if you want them to have independent importance.
2. Redundancy: If you don't want independent control, you can drastically reduce your definition size by grafting your original facade surfaces. You should only need one set of panel controls.
3. Fitness: The problem seems to be too obvious…. It will find that the configuration with the minimum amount of surface will be a flat facade.

For the Ecotect exercise, try setting up independent controls for each face and use Ecotect to find optimal panel configurations.

[http://theprovingground.wikidot.com/local--files/myoboon-hur/ARCH517_FinalProject_Myoboon_20110328.ghx]

However, I have attached a basic method for mapping patterns on surface. In this case, i used voronoi cells. The Watercube is often mistaken for a voronoi pattern.

Get it here.

how do I generate random but-not-really-random points on a 2D surface by using grasshopper components? i am trying to model a facade skin like the Water Cube pattern… i like the points generated by sphere cloud, but its on a 3D surface…

Thanks.

Simon

CamStudio was chosen since it is free and open source.

I have never had a problem with video file corruption… strange.

Please avoid uploading videos directly to the wiki. Space is limited here.

[http://theprovingground.wikidot.com/local--files/myoboon-hur/Final_Progress_myoboon_%2020110328_Tower_Galapagos_SurfaceMinimum.ghx]

Be sure to set the C input on PLine to "True". This will close the polyline….

I also found duplication in your facade vertices. I dunno why you're using the same list of points 3 times; why you're not creating 3 different lists with different vertices. There is an easier way of achieving what you want.

Meet me at school tomorrow if you wanna go over it. I will be there from 1pm-5pm.

-Mohamed

I took a look at your files. However, I would like to know what are you trying to achieve exactly so I can help you. I realized that your "area" component is giving you 112 values. You need to mass add them to end up with "one" numeric value. Prior to "mass adding" them, make sure you flatten the list (you can use "flatten" component).

Let me know if you wanna meet at school tomorrow to get your problem solved.

-Mohamed

Has anyone try out the CamStudio yet? I did that this evening, and would suggest to you to give it a try asap. - Simon

i didn't really have a perfect scenario for the setup. I was thinking if the grid represent a state's borders, and the circles are the railways. if all they all three trains complete their own loop within the same time, at which point on the loop will there be the most people within the borders. (volume of boxes = density of people) I ended up use 1 slider parameter to control all three distance logic systems to restrict an easy-find and multiple solutions. I guess if i were using this to solve the scenario 'problem', I could have made the box into two-dimension geometry and it would run a lot faster. But I was just simply exploring the software…

I am seeing 3 separate distance logic systems being driven by the same slider parameter. You are then finding the top 15 boxes and then asking Galapagos to maximize this final value? But what is the goal?

The reason for the speed issues is that you are generating a very large collection of BReps which are being recomputed for each iteration.

(Update: the solver hit 20 Stagnant after 102 minutes, and didn't solve the problem)

another problem is…there is no F13 key on my mac keyboard…i can't do print screen…

(Update: Chris Norton said from Window Accessories there is an on-screen keyboard which has printscreen button. Thanks Chris!)