Geometry to Pollination - without GH/Dynamo connection

Hello! How is everyone doing? I see a lot of great developments, congratulations!

After some really interesting results, I’m slowly coming back to design performance assessment and I was thinking of ways to use pollination in my workflows.

I have a process that generates designs in simple geometric representations, which do have all info required for simple 3d model construction. I would like to connect those somehow to pollination. Since all this happens outside of any design software, I wanted to ask if it would be possible to translate those to the HB schema and somehow send them to pollination. Are there any simple examples of these perhaps?

Additionally, is an API call to pollination possible at this moment? Could I stream those and receive results in one go?

Thanks in advance! And apologies if some of the above are a bit naive, I’ve been away for too long!

Best,
Theodore.

1 Like

Nevermind the 2nd question, I opened the API docs :smiley:

1 Like

Hi @gabriel_syme ! What is the source for your model geometry?

The input model to most recipes is a file in .hbjson format following the Honeybee schema. There is a python library that can help with the creation of this here.

As you have found, there is an OpenAPI and redoc-documented API to work with for manual integrations. We don’t provide any streaming capabilities right now, unfortunately. But, you should be able to use typical REST conventions (Following the Location header, for instance) to poll results if you need real-time updates.

You can, of course, use API calls to retrieve results after the run is complete, too.

There is a guide here that demonstrates a simple example of an API integration.

Let us know if we can help with anything else!

1 Like

Hi @gabriel_syme! What @tyler said - plus if you are still only working with extruded polygons you might want to use the Dragonfly schema instead of Honeybee schema. That should save you from all the geometry calculations to find the vertices. Dragonfly does that for you.

If you can share a sample code we can help you with translating that to a Dragonfly model which then can be translated to a Honeybee model and you can run it against any of the available recipes.

Hi @mostapha and @tyler ! Thanks for the response, this makes sense. Dragonfly schema would be much better I think yes. By the way, is there an easy way to randomly add windows to the schemas? For example, if I provide a (x,y) coordinate of the center point, the sill height, and window height?

Each layout can have a very simple geometric representations, for e.g.:

'room1': (x1,y1)(x2,y2)...(xn,yn), 'room2': (x1,y1)(x2,y2)...(xn,yn), ....

Of course it is quite easy to adjust this to whatever makes this connection easier. Actually, that translation function I can handle more or less once I understand what the schema expects!

p.s. How can I change my nick name in here? :smiley:

Best,
Theodore.

1 Like

Hi @gabriel_syme,

See the tests here for both your questions. The short answer is here:

I can change it for you. What do you want to change it to?

Looks good! It’s actually really close to what I have already! I’ll give it a shot.

I can change it for you. What do you want to change it to?

Theodoros Galanos :slight_smile:

1 Like

@antoinedao changed it for you. Try to logout and log in back and you should be able to see the change.

1 Like