looking forward to the new releases!!!

Hi Pirouz,

I am going to analyse an urban space with Space Syntax add-on. I imported the line from openstreetmap. They are not lines they are curves, is it makes any issue? Secondly, when I use axial line graph (ALG) component, the add-on gives me a numbers top of some lines(not all of them) which does not make sense. Finaally, would you please guide me how I can get some graphical outcome of this add-on? Generally speaking, Is there any way to get some sense of how this add-on works? I do appreciate your help mate. Thank you

Regards,

Azad

Hi Friends

I've installed the plugin spacesyntax

But it gives error and does not work anymore

When I open the example project still gives error

Please advise your

سلام دوستان

من پلاگین spacesyntax را نصب کردم

اما ارور می ده و کار نمی کنه

مثال پروژه را وقتی باز می کنم باز هم ارور می ده

لطفا راهنمایی کنید

Hey, so I have tried using the space syntax plug in on my program, and its a pretty intensive program with lots of relationships and different spaces. I have done some preliminary connections and used the plug in. Most of the widgets Ive used seem to be working. The analysis widgets are working very well and providing useful insight into the social aspects of the design. BUT the main problem comes when I try to use the ForceDirectedDrawing widget. It poops out and kinda takes my rhino file with it. ANy suggestions? Please help. I have attached some screen shots and the rhino file and gh file.

ERRORSLIDESHOW.pdf

firstProgramSyntax.3dm

FirstProgramsyntax.gh

Hey Edward Whitley, I have the exactly same problem...
Here, the drawing is also locked; I can't draw anything.
You found any Solution?

greets

Hi, let me answer one by one:

@ Azad: you can only work with lines (axial lines crossing each other) I don't think you can immediately convert OSM street centerlines to axial lines. This has a number of probelms, most important of which is methodological. Graph theoretical distance used in Space Syntax measurements (the ones I have implemented) is based on the idea of measuring the [cognitive] steps one takes from one space to another. In architectural configuration these spaces (spatial units) are conventionally considered as convex spaces or let us say rooms. In urban configurations, streets are considered as the [public] space and axial lines are to represent them as the spatial units. In general if you are going to do a large scale urban computation, you can use DepthMap or Depthmapx (the original Space Syntax software application). The main reason I added the Axial-Line Graph component was to use it for education. You cannot yet get visualizations or efficient calculations at urban scale using this one. But please do a bit of research about the methodical problem I mentioned above before going to any other software. This is not a technical problem. 

@ Mahdi:

You have not posted the error message but I guess most probably it's saying something like the method X doesn't exist in Rhino. I have seen the same problem many times... the solution is to make sure your Rhino is updated. You can go to Rhino/Options/Updates to do so. 

@ Edward:

Hi, first of all note that the graph drawing algorithm is not guaranteed to work always, especially if the input graph is non-planar, there is no such proof for any force-directed graph drawing and ours is not an exception. If you are only trying the software with random points and random connections to see if it works then I assure you it won't work for just any combination of points lines and areas. It is simulating a physical system of disks attracted to one another if there should be a link between them. If one disk or coin doesn't fit in between it's neighbors (the problem of area) then of course the algorithm fails because you are trying the impossible. And if the graph is non-planar per se, which is your case, then again it fails. It of course happens in real-world architectural bubble diagrams that they don't constitute a planar graph, simply because there are many buildings that can only be built with the same abstract configuration in more than one floor. But in this case, you can divide the graph into planar components and proceed. I believe We have mentioned these points in the help text of the component itself. Please have a look.  

@ Ortler:

Please read my answer to Edward; most probably the same issue. 

Hey Pirouz Nourian,

Thanks for the fast reply.
Im not the Graph-Expert but I guess that my Graph is planar(topological).
No Loops at all, but some intersections of the (real/drawn) edges... 

Without Loops, shouldn't it possible to remap it?

 
I get the point of your post addressed to Edward, but I do not understand why the component fails in my case.

Can u have a look at it and tell me what I can do to get it work?

greets Mark

FlaechenAnalyse.zip

Hi Ortler, You're welcome. I just checked your file; I understand it is a case of analysis of an already existing plan. You need a nice graph drawing if I am not mistaken. Right? Even if the graph is planar, still you need to have a combination of area values to make a kissing-disks or coin-graph drawing. In some nodes you have too many connections to a node that means for the disks to kiss each other that node must be significantly bigger than all its neighbors. Similar problems might arise when one big coin is supposed to be trapped in the middle of a bunch of small coins. I understand this might be frustrating because you already see a feasible planar drawing of this graph as your plan. But note that for circles to meet each other in this way these challenges exist. In general you need to provide a set of area values to make the kissing disk drawing possible. In order to do this, I would either break the graph in parts and approach one by one or try to lower the repulsion force and then put numbers one by one such that their combination would produce a feasible planar configuration of coins. No easy way in either case. Having said all this, please note that if you don't manage to draw a graph with this component it doesn't mean that you cannot use analysis components. I started recognizing the points that connect to many other and giving them a higher area; lowering the repulsion and limiting the iterations at the same time, this way you can produce some acceptable bubble drawings and proceed easier -perhaps. Now they are closing the faculty building and kicking everyone out... FlaechenAnalyse_PZN_Ed.gh

Hey, so I have deleted a couple connections in the program, and I have manipulated the attraction and repulsion inputs, and I have seemed to solve the problem... unfortunately it is still a little touch and go if I move those sliders too much. But ultimately Im getting a pretty great output I think. Thank you so much Pirouz, for the help and the program. 

Here is my new file:

2ndProgramSyntax.3dm

2Programsyntax.gh

Hi Edward, happy to hear that. Looks like you have pushed it to the limit :)

Still a few coins don't fit around their hubs, but yes if you let them relax by lowering repulsion you can get a result. Good luck with the rest. 

Hi Pirouz

First of all thank you very much for the plugin. I installed and used it in another computer and it was fine but in mine it gives me this error...

An error occured during GHA assembly loading:
Path: C:\Users\Frane\AppData\Roaming\Grasshopper\Libraries\SpaceSyntax_for_GenerativeDesign.gha
―――――――――――――――――――――――――――――――――――――――――――――――――――――――――――
The GHA file is either damaged or has been made with
a newer version of .NET than Grasshopper itself.
You can try toggling the "Memory Load" option in the
GrasshopperDeveloperSettings command window, it might help.
――――――――――――――――――――――――――――――――――――

Any idea why?

Thanks again

Hi Pirouz

I feel stupid... I thought I tried toggling "Memory Load" and it didn't work... I did it now and it works. 

So thanks anyway

Bye

Hi Frane, 

Happy to hear that it works now; but could you please tell me how exactly you opened this developer settings window?

Best,

Pirouz

Hi Pirouz.

It works but I have to do it every time I start GH.

You can open the window with the "GrasshopperDevelopperSettings" command in Rhino. 

Hi Frane, reading the error message again; I get the impression that your GH is not up to date... correct?

Hi Pirouz

Strange! Anyway I am planning to release a new version with a small bug fix soon; in this version I will take care of a few issues related to the .Net and Rhinocommon version. Hopefully it will be solved. I need to think about it more... 

Hi,

Pirouz thanks so much for all the work. Do you think you can upload the new version, that you refered earlier, the space syntax layout with cardinal directions? Thanks

Dear Pirouz,

i would like to ask a suggestion to you beacuse is impossible to figure out the error of my Diskograph component.

basically I have a distributiion with 29 points and connections among them. I tested a definition with 4 points and working fine, but adding more points the NGV output is giving a strange codification of the point coordinates.

I will be very glad to hear from you.

Thanks in advance

Dario

error.jpg 

Hi Dario, just responded via email.  

Hi Tata, thanks for your kind words. I just need a few hours free to finalize a new version. This week is quite hectic, hopefully next week I will share the new version. 

Hi Pirouz! Any news on the new version?

Hi Tata, I made some last changes to the latest version on Sunday, it's basically ready but I need a few more hours to rebuild the assembly and publish. Soon...

Hi, it says download link is down ..

Hi all, the new version is now available at: https://sites.google.com/site/pirouznourian/syntactic-design?pli=1

There might be a little issue with the visualization of disks (they might have been turned into dark colors); haven't figured out yet what it is. But there is a simple solution for the time being: zoom on your work from your perspective view-port and change the name of this view port to TOP.

Apart from this, make sure you read the notes on the new components. One of them is just a dummy place holder. 

Big Thanks Prof Nourian, its really very useful, and am one of those who is waiting it.

Really i appreciate your Time & Effort.

Thanks Rafat for your compliments. I would love to be a professor but I'm just "a simple kid with a simple passion for simple things" :) technically still a PhD researcher/instructor!  Hope the new version is running smoothly. 

Hi Pirouz, I'm using Space Syntax to configure the functional layout in an urban design project. 
I've noticed that with a quite high range of functions and especially when all this elements have too many interconnections between them, it became really hard to obtain a clear configuration. For ex in my case I've got 13 functions with more than 30 connections, but the result is incomprehensible, as you can see from the image.

I've tried to search a previous conversation about this issue, but I didn't found a solution. Am I making some mistakes or do I pretend to use too many imputs? Do you have any suggestion?

Thank you!!

Hallo Leonardo,

I might not be able to suggest anything but I would like to point out two things:

Syntactic is using force-directed graph drawing (as fare as I know). This comes with certain restraint:

http://en.wikipedia.org/wiki/Force-directed_graph_drawing

Secondly it might be that you graph ain't planar:

http://en.wikipedia.org/wiki/Planarity_testing

Maybe this helps.

Hi Richard, thanks for the tips.. I've checked and all the points and lines are planar.. 

So, does anyone has suggestions on how fix this issue?

Hello Leonard,

Ain't nothing to do with the lines being planar, please read the postet articles.

ok.. I read the two links.. but I still cannot figure out if I'm wrong and where.. for ex, about the planarity, I think that it's just impossibile for me to avoid the edges intersection.

Well as said explained in the article about force directed graph drawing, it is an iterative process so one thing you can do is to increase the number of iterations (It_#). Also you can play with the strength of Attraction / Repulsion (St_Atr / St_Rep).

However if the graph is not planar than there is a conflict, since an edge means attraction... . 

@ Richard: Many thanks for your explanations Richard! 

@ Leonardo: As Richard says, if the graph is non-planar then you cannot expect a kissing-disk drawing to be created without crossings. If you want to get a result that is acceptable, you can lower the strength of the springs (ST-Atr) as Richard has suggested. Note that this words of yours "it's just impossible for me to avoid the edges intersection" mean exactly that the graph is non-planar. 

Tank you very much Pirouz and Richard.. now everything seems more clear!!

Hi, you need to drag the GHA file from the folder installed by the installer and drop it on your GH canvas. Note that the installer asks for the location of this folder during installation. If you don't specify a location it puts the folder in a default location. Please read the notes in the folder. 

very nice work thank you so much ! stay blessed:)

Hello, 

I am really enjoying working with this component, and I am wondering if it is possible to add some parameters and constraints related to energetic and climate contexts.... like p.e. parameters related to sun path (so i will be sure that my graph and relations will be associated to the site conditions...).. so in other words, the functions will be in the right place...

Thank you very much for this work also 

Hi, 

@ Arrow Morphow: thanks for your kind message!

@  dalec: I have not found a generic stable way of doing that in 3D yet. Let me clarify something: the tools for plan layout do not perform any optimization whatsoever. The dual graph component can reveal possible 'topological embedding's for a planar graph that describes desired spatial connectivity links. It does that by introducing adjacency links that were not mentioned in the initial graph. Simply put, it shows multiple ways in which rooms can be laid out on a plane so as to allow for the mentioned spatial connections.If the input graph is non-planar (i.e. you cannot draw it on a piece of paper without edges crossing one another) then this process fails. Speaking of optimization; you need to have a notion of goodness interpreted as a measurable 'objective function' that you can seek to maximize (or minimize). There is no such thing in the toolkit at the moment. As for performance measures, the idea is that the designer initially puts in the desired adjacency such as "kitchen facing northeast"  having in mind lighting conditions and alike. I can imagine that such performance criteria can be integrated but have no idea how generic they can be. As for a toolkit we only focus on tools that are generic-purpose. But of course I would be happy to hear your ideas on this subject; perhaps we can find a way...

Hi Waldemar Felipe, 

Happy to hear that the tools can be useful in social housing design! 

As I have written below, the 'dual graphs' that look like rooms are only showing the possibilities for laying out rooms on a plane such that they allow for the desired spatial connections put in by the user. There are ways to draw a rectangular dissection with the same topological properties. We did not manage to make such drawings; that is why I have marked these tools as funny/dummy/place holder:

I would like to develop these further; but unfortunately I don't have the time for that. Please read the message in the example file. If anybody can develop these tools as I have explained, I will include them in the next release. Right now, I am thinking of an alternative way to grow spaces into each other, a completely different story that starts from the dual graph of rooms...

Hello Pirouz, and thank you for your answer. 

I know how your toolkit function, and I know that it doesn't provide an optimization, but I think if we find a way to let him provide an optimization by adding multicriteria constraints It would be great...,

I am using this process of generic and parametric modeling in a research project for my thesis. 

It seems that architecture programming is being discussed more and more. Optimization of the connectivity between spaces of an architectural projects is needed specially for big projects. 

Therefor, I wanted to to test the connectivity between spaces not only in 2D level, but with more than one level, (referring to the different stories that we can in the building). Since some of the spaces connected to each other in a graph (2D level) might have to be connected with other spaces in the next level.. I would appreciate also if we can discuss this issue, and having a feedback from you of what you think and what can you propose as a method of solutions.

Hi dalec, speaking of 3D configurations, there is nothing limiting the node-link graph to model a 3D configuration as a graph (to begin with). Once you have your configuration modeled as a graph you can of course do all the analyses of Space Syntax too. The graph is however different from its embedding (drawing). How you put a bunch of points in space referring to the nodes (representing spaces) is a matter of graph drawing. I can do something similar to disk drawing with spheres but I don't see how useful it could be. An approach similar to Tutte drawing (as in the new components) is already quite complicated in 2D; I do not have a clear idea as to how that process can be generalized to 3D; but I have not given up. I am thinking of alternatives. Might show some alternative ways in future here. I would also welcome ideas from everyone in the group. 

Regarding optimization; we need to define objective functions. What kind of objective functions do you have in mind? 

Re 3D: Richard Schaffranek has some very interesting 3D configuration experiments, I recommend that you try his toolkit SpiderWeb and have a look at this workshop notes: http://www.gbl.tuwien.ac.at/Archiv/digital.html?name=AAG2014

Hello and thank you for your replies. 

I will take a look of these links and get back to you.

Re, can I find other examples or tutorial for Space Syntax using other components? 

Thank you