lly it should not make much of a difference - random number generation is not affected, mutation also is not. crossover is a bit more tricky, I use Simulated Binary Crossover (SBX-20) which was introduced already in 1194:
Deb K., Agrawal R. B.: Simulated Binary Crossover for Continuous Search Space, inIITK/ME/SMD-94027, Convenor, Technical Reports, Indian Institue of Technology, Kanpur, India,November 1994
Abst ract. The success of binary-coded gene t ic algorithms (GA s) inproblems having discrete sear ch sp ace largely depends on the codingused to represent the prob lem variables and on the crossover ope ratorthat propagates buildin g blocks from pare nt strings to childrenst rings . In solving optimization problems having continuous searchspace, binary-co ded GAs discr et ize the search space by using a codingof the problem var iables in binary st rings. However , t he coding of realvaluedvari ables in finit e-length st rings causes a number of difficulties:inability to achieve arbit rary pr ecision in the obtained solution , fixedmapping of problem var iab les, inh eren t Hamming cliff problem associatedwit h binary coding, and processing of Holland 's schemata incont inuous search space. Although a number of real-coded GAs aredevelop ed to solve optimization problems having a cont inuous searchspace, the search powers of these crossover operators are not adequate .In t his paper , t he search power of a crossover operator is defined int erms of the probability of creating an arbitrary child solut ion froma given pair of parent solutions . Motivated by t he success of binarycodedGAs in discret e search space problems , we develop a real-codedcrossover (which we call the simulated binar y crossover , or SBX) operatorwhose search power is similar to that of the single-point crossoverused in binary-coded GAs . Simulation results on a number of realvaluedt est problems of varying difficulty and dimensionality suggestt hat the real-cod ed GAs with t he SBX operator ar e ab le to perform asgood or bet t er than binary-cod ed GAs wit h t he single-po int crossover.SBX is found to be particularly useful in problems having mult ip le optimalsolutions with a narrow global basin an d in prob lems where thelower and upper bo unds of the global optimum are not known a priori.Further , a simulation on a two-var iable blocked function showsthat the real-coded GA with SBX work s as suggested by Goldberg
and in most cases t he performance of real-coded GA with SBX is similarto that of binary GAs with a single-point crossover. Based onth ese encouraging results, this paper suggests a number of extensionsto the present study.
7. ConclusionsIn this paper, a real-coded crossover operator has been develop ed bas ed ont he search characte rist ics of a single-point crossover used in binary -codedGAs. In ord er to define the search power of a crossover operator, a spreadfactor has been introduced as the ratio of the absolute differences of thechildren points to that of the parent points. Thereaft er , the probabilityof creat ing a child point for two given parent points has been derived forthe single-point crossover. Motivat ed by the success of binary-coded GAsin problems wit h discrete sear ch space, a simul ated bin ary crossover (SBX)operator has been develop ed to solve problems having cont inuous searchspace. The SBX operator has search power similar to that of the single-po intcrossover.On a number of t est fun ctions, including De Jong's five te st fun ct ions, ithas been found that real-coded GAs with the SBX operator can overcome anumb er of difficult ies inherent with binary-coded GAs in solving cont inuoussearch space problems-Hamming cliff problem, arbitrary pr ecision problem,and fixed mapped coding problem. In the comparison of real-coded GAs wit ha SBX operator and binary-coded GAs with a single-point crossover ope rat or ,it has been observed that the performance of the former is better than thelatt er on continuous functions and the performance of the former is similarto the lat ter in solving discret e and difficult functions. In comparison withanother real-coded crossover operator (i.e. , BLX-0 .5) suggested elsewhere ,SBX performs better in difficult test functions. It has also been observedthat SBX is particularly useful in problems where the bounds of the optimum
point is not known a priori and wher e there are multi ple optima, of whichone is global.Real-coded GAs wit h t he SBX op erator have also been tried in solvinga two-variab le blocked function (the concept of blocked fun ctions was introducedin [10]). Blocked fun ct ions are difficult for real-coded GAs , becauselocal optimal points block t he progress of search to continue towards t heglobal optimal point . The simulat ion results on t he two-var iable blockedfunction have shown that in most occasions , the sea rch proceeds the way aspr edicted in [10]. Most importantly, it has been observed that the real-codedGAs wit h SBX work similar to that of t he binary-coded GAs wit h single-pointcrossover in overcoming t he barrier of the local peaks and converging to t heglobal bas in. However , it is premature to conclude whether real-coded GAswit h SBX op erator can overcome t he local barriers in higher-dimensionalblocked fun ct ions.These results are encour aging and suggest avenues for further research.Because the SBX ope rat or uses a probability distribut ion for choosing a childpo int , the real-coded GAs wit h SBX are one st ep ahead of the binary-codedGAs in te rms of ach ieving a convergence proof for GAs. With a direct probabilist ic relationship between children and parent points used in t his paper,cues from t he clas sical stochast ic optimization methods can be borrowed toachieve a convergence proof of GAs , or a much closer tie between the classicaloptimization methods and GAs is on t he horizon.
In short, according to the authors my SBX operator using real gene values is as good as older ones specially designed for discrete searches, and better in continuous searches. SBX as far as i know meanwhile is a standard general crossover operator.
But:
- there might be better ones out there i just havent seen yet. please tell me.
- besides tournament selection and mutation, crossover is just one part of the breeding pipeline. also there is the elite management for MOEA which is AT LEAST as important as the breeding itself.
- depending on the problem, there are almost always better specific ways of how to code the mutation and the crossover operators. but octopus is meant to keep it general for the moment - maybe there's a way for an interface to code those things yourself..!?
2) elite size = SPEA-2 archive size, yes. the rate depends on your convergence behaviour i would say. i usually start off with at least half the size of the population, but mostly the same size (as it is hard-coded in the new version, i just realize) is big enough.
4) the non-dominated front is always put into the archive first. if the archive size is exceeded, the least important individual (the significant strategy in SPEA-2) are truncated one by one until the size is reached. if it is smaller, the fittest dominated individuals are put into the elite. the latter happens in the beginning of the run, when the front wasn't discovered well yet.
3) yes it is. this is a custom implementation i figured out myself. however i'm close to have the HypE algorithm working in the new version, which natively has got the possibility to articulate perference relations on sets of solutions.
…
he example file to this file so you can give it a try with any version of Honeybee that you're already using. The only requirement is to have OpenStudio installed as the component is using OpenStudio libraries to parse gbXML files. If you're using the latest version available on github the component is also available under WIP tab.
Why?
The main purpose of developing this component is to save time and effort for importing Revit models for energy and daylight analysis. It bothers me to see a lot of smart people spend a lot of time to just come up with solutions just to get the geometry from Revit to Honeybee for analysis. This component is not solving all the issue but is a first step forward. In an ideal world, the future version of Honeybee, which works both under DynamoBIM and Grasshopper should address this issue but that can take some time to be fully ready!
How?
To use this component you need to Export your Revit model as gbXML and then use the file path to load the file into Grasshopper. There are several resources available online on how to prepare the analytical model in Revit and export the gbXML file. Here is an image for importing the Revit 2017 sample model using the default settings. As you can see the model will be just as good as what your original gbXML file from Revit is.
What can be improved?
Well, there are several items that can be improved and they are mostly not on us. To get it started I add what I think are the 3 main shortcomings and my thoughts on how they can be addressed in the future. Feel free to add what you think needs to be added to this list in the comments section.
1. Revit analytical models and as the results gbXML files, by design, are not intended to be clean. Watch this presentation from the Autodesk University to see the logic behind this approach which in short is it doesn't matter for a large scale early stage energy model. Well, This will be quite a problem for studies that you can do with Honeybee. Included but not limited to daylight and comfort analysis.
The best solution that I can think of, until Autodesk fixes their exporter, is to use Revit Rooms and Spaces and generate a clean model from the scratch. We have already tried this approach in Revit but since the Revit API doesn't provide access to Room openings we had a very hard time to get it to work.
That's why that I opened an idea on Revit ideas to get over this issue. With your support we already have 81 votes, but it hasn't been enough to make them to consider the idea for an official review. If you haven't voted already and you think this will be a helpful feature take a moment and vote so we can have it implemented at some point in the future.
2. There is no way (that I know) to export only part of the model. The way export gbXML is set up in Revit is to export the whole model once together. As a result, if you have a huge model with 100 rooms and you want to get one of the rooms into Honeybee using this component you have to export the whole model, which can take some time, and then import them all back into Grasshopper. To partially address this issue I added an input to the component that allows you input a list of names for rooms that you're interested to be loaded into Grasshopper. You can use the name of the room/space in Revit as an input for the component.
3. The component doesn't import adjacencies, loads, schedules and HVAC systems. I wasn't able to export a gbXML file from Revit with any of this data except for the adjacency, but even if you can do that, the component currently can only import geometries and constructions. I hope we get access to 1 and so we don't have to use the xml file approach at all, but if that takes a very long time then we will add these features to the component.
Happy 2017!
Mostapha…
it seems that was this. Now all is working fine !
Glad that it worked! But I am still a bit worried. Gismo components only modify the gdal-data/osmconf.ini file and no other MapWinGIS file. So your MapWinGIS installation files should not be compromised. The fact that you did not get the "COM CLSID" error message when running the "Gismo Gismo" component suggests that MapWinGIS has been properly installed. So I wonder if the cause for the permanent "invalid shapes" warning has again something with the fact that your system is again not allowing the MapWinGIS to properly edit the osmconf.ini. Maybe this problem will appear again, and again, and reinstallation of MapWinGIS every time can be somewhat bothersome.
- About the terrain generation, is it possible to have the texture from google or other provider mapped onto the terrain surface from gismo component ? (Same as using the ladybug terrain generator in fact). I try to used the image extracted by ladybug component and then applied it to the gismo terrain but the texture is rotated by 90°.
The issue with the rotation can be solved by swapping/reversing the U,V directions of the terrain surface. A slightly more important issue is that terrain surface generated with Gismo "Terrain Generator" component might have a bit smaller radius than what the radius_ input required. This stems from the fact that the terrain data first needs to be downloaded in geographic coordinate system, and then projected. Some projecting issues may occur at the very edges of the projected terrain, so I had to slightly cut out the very edges of the terrain which results in the actual terrain diameters being slightly shorted in both directions. This means that if you apply the same satellite image from Ladybug "Terrain Generator" component to Gismo "Terrain Generator" component the results may not be the same.I attached below a python component which tries to solve this issue by extending the edges of Gismo "Terrain Generator" terrain, and then cutting them with the cuboid of the exact dimensions as the radius_ input. Have in mind that this extension of the original terrain at its edges is not a correct representation of the actual terrain in that location. But rather just an extension of the isoparameteric curve of the terrain surface. So basically: some 0 to 10% (0 to 10 percent of the width and length) of the terrain around all four edges is not the actual terrain for that location, but rather just its extension.The python component is located at the very right of the definition attached below.
Also, if you would like to use the satellite images from Ladybug "Terrain Generator" component along with "OSM shapes", sometimes you may find slight differences in position of the shapes. This is due to openstreetmap data not being based on Google Maps (that's what Ladybug "Terrain Generator" component is using), but rather on Bing, MapQuest and a few others.
- About the requiredKeys_ input of OSM shapes, I understand what you mean and your advice, but in most cases I use it, the component was working fine even without input. I think it's better to extract all tags, values and keys of the selected area, instead of searching for specific ones as I try to find all data related to what I want after, isn't it ? To check what keys are present on the area also.
Ineed, you are correct.I though you were trying to only create a terrain, 3d buildings and maybe find some school or similar 3d building, for these two locations. The recommendation I mentioned previously is due to shapefiles having a limit (2044) to how many keys it can contain. This requires further testing of some big cities locations with maybe larger radii, which I haven't performed due to my poor PC configuration. But in theory, I imagine that it may happen that a downloaded .osm file may have more than 2044 keys. In that case shapefile will only record 2044 of them, and disregard the others. That was my point.But again 2044 is a lot of keys, and I haven't been checking much this in practice. For example, when I set the radius_ to 1000 meters, and use your "3 Rue de Bretonvilliers Paris" location I get around 350 something keys, which is way below the 2044.Another reason why one should use the requiredKeys_ input is to make the Gismo OSM components run quicker: for example, the upper mentioned 350 something keys will result in 350 values for each branch of the "OSM shapes" component's "values" output.Which means if you have 10 000 shapes, the "OSM shapes" component will have 10 000 branches with 350 items on each branch (values). This can make all Gismo OSM components very heavy, and significantly elongate the calculation process.With requiredKeys_ input you may end up with only a couple of tens of items per each branch.Sorry for the long reply.…
Added by djordje to Gismo at 8:57am on June 11, 2017
This blog post is a rough approximation of the lecture I gave at the AAG10 conference in Vienna on September 21st 2010. Naturally it will be quite a different experience as the medium is quite…
Added by David Rutten at 3:27pm on September 24, 2010
t. So here we go!
1. Honeybee is brown and not yellow [stupid!]...
As you probably remember Honeybee logo was initially yellow because of my ignorance about Honeybees. With the help of our Honeybee expert, Michalina, now the color is corrected. I promised her to update everyone about this. Below are photos of her working on the honeybee logo and the results of her study.
If you think I'm exaggerating by calling her a honeybee expert you better watch this video:
Thank you Michalina for the great work! :). I corrected the colors. No yellow anymore. The only yellow arrows represent sun rays and not the honeybee!
2. Yellow or brown, W[here]TH Honeybee is?
I know. It has been a long time after I posted the initial video and it is not fun at all to wait for a long time. Here is the good news. If you are following the Facebook page you probably now that the Daylighting components are almost ready.
Couple of friends from Grasshopper community and RADIANCE community has been helping me with testing/debugging the components. I still think/hope to release the daylighting components at some point in January before Ladybug gets one year old.
There have been multiple changes. I finally feel that the current version of Honeybee is simple enough for non-expert users to start running initial studies and flexible enough for advanced users to run advanced studies. I will post a video soon and walk you through different components.
I think I still need more time to modify the energy simulation components so they are not going to be part of the next release. Unfortunately, there are so many ways to set up and run a wrong energy simulation and I really don’t want to add one new GIGO app to the world of simulation. We already have enough of that. Moreover I’m still not quite happy with the workflow. Please bear with me for few more months and then we can all celebrate!
I recently tested the idea of connecting Grasshopper to OpenStudio by using OpenStudio API successfully. If nothing else, I really want to release the EnergyPlus components so I can concentrate on Grasshopper > OpenStudio development which I personally think is the best approach.
3. What about wind analysis?
I have been asked multiple times that if Ladybug will have a component for wind study. The short answer is YES! I have been working with EFRI-PULSE project during the last year to develop a free and open source web-based CFD simulation platform for outdoor analysis.
We had a very good progress so far and our rockstar Stefan recently presented the results of the work at the American Physical Society’s 66th annual DFD meeting and the results looks pretty convincing in comparison to measured data. Here is an image from the presentation. All the credits go to Stefan Gracik and EFRI-PULSE project.
The project will go live at some point next year and after that I will release the Butterfly which will let you prepare the model for the CFD simulation and send it to EFRI-PULSE project. I haven’t tried to run the simulations locally yet but I’m considering that as a further development. Here is how the component and the logo looks like right now.
4. Teaching resources
It has been almost 11 months from the first public release of Ladybug. I know that I didn't do a good job in providing enough tutorials/teaching materials and I know that I won’t be able to put something comprehensive together soon.
Fortunately, ladybug has been flying in multiple schools during the last year. Several design, engineering and consultant firms are using it and it has been thought in several workshops. As I checked with multiple of you, almost everyone told me that they will be happy to share their teaching materials; hence I started the teaching resources page. Please share your materials on the page. They can be in any format and any language. Thanks in advance!
I hope you enjoyed/are enjoying/will enjoy the longest night of the year. Happy Yalda!
Cheers,
-Mostapha
…
DP ($$$ aside), GC, and Grasshopper. Arthur’s original question is very important
and the exact question (and hopefully answer) I was hoping to find on a
forum.
“How to take intelligent 3D parametric generative design models (scripting, etc.) into 2D documents?" Or, deliver the 3D design for evaluation, bid, construction, etc.
I am intrigued by Jon’s comments in the same thread and would like to know how I can learn more about the process (and
pitfalls) of turning over a 3D digital generative models to a contractor/fabricator.
Are there any industry guidelines established I could use as a reference to guide our firm through this type of uncharted territory?
Arthur’s question is very reminiscent of 10 years ago when I was frustrated with the amount of time spent on the development of a 3D model design (physical and/or virtual) only to have to wipe the table clean and start the process all over again in 2D in order to document the project for delivery. From this I jumped head first into BIM and Revit, vowing never to go back to unintelligent 2D line work. I am now working on Bentley software (v8i: Microstation and Bentley Architecture) with the access and desire to venture into Generative Components. I am very intrigued by Rhino/Grasshopper primarily with the apparent ease of use and available resources assisting in the learning process – something not really available with Bentley.
In hindsight, as I am doing my software research I think the current use of Revit and BA (Bentley Architecture) are more of a “bridge”
between the past (decades of digital 2D work, i.e. AutoCAD) and where hopefully
we all will be someday in the near future (100% 3D modeling, i.e. Digital
Project??). Without having the experience
it would appear that DP/CATIA (PLM software) are closer to this than any other
type of software. As complicated as the
industry standards are for the automobile and airline industry, I feel we
(architectural industry and others) are heading in a similar direction with
total understanding (PLM/ Evidence Based Design) of a design (a whole other topic). If anything I think the market will begin to
demand it sooner or later.
Gehry (DP) article NY Times:
http://www.nytimes.com/2009/02/11/business/11gehry.html
I know these type of broad discussions (software vs. software) can be blown out of proportion on forums, but I am would like to read
the pulse of those who are already in the trenches (using Grasshopper, CATIA, Digital Project, Generative Components, others??) and hear your thoughts. Just as valuable would be other threads,
industry articles/reviews of 3D parametric generative design software.
Thanks,
Boyd…
ting at multiple geometries in the same location. I simply sorted the list of values and used the Delete Consecutive component. This potentially rearranges the order of values but I don't think that matters in your case. I also threw in an Int component which actually seems to make a difference (try sidestepping it and you will see!).
2-I flattened the output of the mesh component before sending it to union. This ensures that the original mesh is booleaned once with all the components rather than individually with each of the 86 components.
Is this what the result should look like?
One suggestion for future postings: when referencing geometry in rhino, it often helps if you attach your rhino file as well so people don't have to guess where you are starting from.
If you have further questions, just ask ;-)
cbass…