Hello 
community!
We have improved the comfort map recipes to reduce file sizes by 50-70%, enhance performance with NumPy, and speed up execution on Pollination Cloud, saving you both storage and CPU hours.
These optimizations apply to all the three comfort map recipes:
- pmv-comfort-map
- adaptive-comfort-map
- utci-comfort-map
What’s new?
Here is the list of the most important improvements to the recipes.
Reduced file size
We have optimized file storage by saving Radiance files in floating point precision and converting timestep-based files into NumPy files. Depending on the model, this change reduces the results folder size between 50-70%.
Enhanced performance with NumPy
NumPy has been integrated into more calculation steps within the ladybug-comfort library. This contributes to the smaller file sizes mentioned above and enhances the computational efficiency of certain tasks. While some steps still do not utilize NumPy, this update lays the groundwork for future optimizations.
Faster execution on Pollination cloud
The recipes have been restructured to run faster on Pollination Cloud. This performance boost is specific to cloud-based runs, ensuring quicker results when utilizing Pollination’s cloud infrastructure. The improvement affects the duration of the simulation, with the most significant impact seen in reduced CPU hours. Combined with the reduced file sizes, these improvements will save you both compute time and data storage on Pollination Cloud.
So what?
After these improvements, you can now run larger models in a shorter time without making any changes to your models. For instance, before making these improvements, this model needed several considerations to run successfully but now it runs with no issues out of the box.
You can check out the sample runs for each recipe on Pollination:
Let us know if you have any questions.
