topographic streamlines creating fissures

Replies to This Discussion

Permalink Reply by peter fotiadis on November 7, 2015 at 12:02am

Hmm ... appears "easy" but .. in fact it's not.

You mean that given a 3d terrain with some data (curves, points, cats, dogs, whatever) you need to define (*) "regions" of interest according a variety of scenarios (for instance: find bridges and banks without top security and some ...suitable escape routes, he he)  and then connect them by curves laying (or not) to that terrain? (creating some kind of histogram). Or maybe perform some "relation" action (say: a KMeans cluster analysis or HAC of other) and make the histogram(s) due to that?

(*) if this includes "on-the-fly" picks and quite obviously recording history of actions is done solely via code (NOT stuff for a novice by ANY means).

Moral: Another giant pile of worms that one.

Q: Is this Academic?

Permalink Reply by Denise Luna on November 8, 2015 at 5:38pm

This sort of academic, it is independent research. 

How may I do this at the moment without integrating data or coding? Sort of a proposal of how our work may look like... 

Thank you for your help Peter!

Permalink Reply by peter fotiadis on November 9, 2015 at 12:16am

Well ... I don't think that's possible without code (and some parts of it are a bit tricky) at least dealing with things the way I see it.

For instance:

1. Assume that we have these "topics" of interest sampled into data trees (if not we must "add" them "manually" == code: get this item from Rhino and put it there) into collections.

2. Then we must perform some kind of selection(s) on a per individual item basis and THAT is in 99% of cases "manual" (== code) or on a per "global basis" (hard or soft clusters et all == code). If clusters are hierarchical and some kind of dendrogram is required ... this obviously means ... er ... more code.

3. Doing the 2 we use some kind of input by means of sliders (say pairs of 2: for branches and items) and therefor MAY their values cause slider control issues (== code). For instance IF this slider yields a x event > do this and that to some other sliders.

4. Then perform the "histogram" required and obviously treat this as just a variant (i.e. a possible solution out of a given collection witch is variable) meaning ways to "store" this into parameter(s) (as persistent data). This also requires code.

In a nutshell (and oversimplified): given a collection of "shapes" pick some make the histogram, store the result (or do something with that and store the outcome as well) recall some other for any reason, modify it, stored it ... and then repeat until the end of time (or worst: until you are out of espresso).

As I said: NOT a task for a novice AND NOT a task for someone not familiar with code matters (But I guess that you qualify in both areas, he he).

I do this type of things day in day out (but for real-life AEC purposes) therefor I could make a "simple demo" (add some "" more) but ... well ... you are warned, he he

But in case that you take the wrong decision (you are warned) we must use Skype a bit.

Permalink Reply by Kim hauer on November 7, 2015 at 2:30pm

Presumably this city or town is already in existence, somewhere on the planet. Most Architects and planners now use Google maps which can give them all the information they need. In fact SketchUp which used to be owned by Google has a very good interface that allows designers to insert their design proposals into existing terrain  https://www.youtube.com/watch?v=-E2QwL467cI

I don't see how Grasshopper is going to help you with this task!

Permalink Reply by Denise Luna on November 8, 2015 at 5:11pm

Thanks for the reply Kim!

Maybe I didn't explain myself very well, what I want to do it's to create a new structure above this city, I already have the topography, but I want this structure to be according to that. 

And this structure will be like my sketches, that's the part I want to do in grasshopper.