# Predictor Data Shapefile units from Recreation Model

**URL:** https://community.naturalcapitalalliance.org/t/predictor-data-shapefile-units-from-recreation-model/2521
**Category:** Software Support
**Tags:** recreation
**Created:** [February 22, 2022, 7:31pm UTC](https://community.naturalcapitalalliance.org/t/predictor-data-shapefile-units-from-recreation-model/2521 "2022-02-22T19:31:34Z")
**Posts on this page:** 4
**Page:** 1

<div class="post-metadata">

### Author: ![jam17m](https://avatars.discourse-cdn.com/v4/letter/j/35a633/32.png) [@jam17m](https://community.naturalcapitalalliance.org/u/jam17m)
#### Post date: [February 22, 2022, 7:31pm UTC](https://community.naturalcapitalalliance.org/t/predictor-data-shapefile-units-from-recreation-model/2521/1 "2022-02-22T19:31:35Z")

</div>

Hello!

I was hoping to get some help with understanding the contents of the “predictor\_data.shp” included in the model output for Recreation and Tourism Model.

I am interested in modeling PUD visitation rates associated with coastal habitats like seagrass beds and coral reefs in Florida, controlling for other relevant socioeconomic factors. I’ve included several potential predictors including the percent overlap of these marine foundation habitats using the “polygon\_area\_coverage” option. My understanding is that the values should range somewhere between 0 and 100% depending on how much a given habitat overlaps the pixel.

But in the resulting predictor\_data shapefile, it looks like these values have much different ranges.

 ![image](https://canada1.discourse-cdn.com/flex031/uploads/naturalcapitalproject/original/2X/3/3de2b0f09de74c2f1741bc02cc5f57375642ca8d.png)

Are the units for habitat overlap actually something different? Is there some kind of scaling that happens that I’m not aware of? Or does the “predictor\_data.shp” not actually represent what I think it does (i.e,. the cell values for each calculated predictor)?

Thanks so much for your help!

Sincerely,  
Jennifer

PhD Candidate  
FSU Geography  
jam17m@my.fsu.edu

---

<div class="post-metadata">

### Author: ![dave](https://yyz1.discourse-cdn.com/flex031/user_avatar/community.naturalcapitalalliance.org/dave/32/70_2.png) [@dave](https://community.naturalcapitalalliance.org/u/dave)
#### Post date: [February 22, 2022, 8:14pm UTC](https://community.naturalcapitalalliance.org/t/predictor-data-shapefile-units-from-recreation-model/2521/2 "2022-02-22T20:14:35Z")

</div>

Hi @jam17m , thanks for posting.

It sounds like you wanted to use `polygon_percent_coverage` as the predictor type. `polygon_area_coverage` calculates the actual area (maybe square meters in your case?) of the predictor in each grid cell. So those are the values you’re reporting in your summary stats. Switching that `type` in the predictor CSV should give you the percentage values you’re expecting.

---

<div class="post-metadata">

### Author: ![jam17m](https://avatars.discourse-cdn.com/v4/letter/j/35a633/32.png) [@jam17m](https://community.naturalcapitalalliance.org/u/jam17m)
#### Post date: [February 23, 2022, 2:17pm UTC](https://community.naturalcapitalalliance.org/t/predictor-data-shapefile-units-from-recreation-model/2521/3 "2022-02-23T14:17:25Z")

</div>

Ah, yes! I see now I had just taken the wrong command to use in the predictors table.

Thank you for your help!

J

---

<div class="post-metadata">

### Author: ![system](https://canada1.discourse-cdn.com/flex031/uploads/naturalcapitalproject/original/2X/8/8d95591aa074da87ea0fdd61ca47c1c3c0588aa7.png) [@system](https://community.naturalcapitalalliance.org/u/system)
#### Post date: [March 2, 2022, 2:18pm UTC](https://community.naturalcapitalalliance.org/t/predictor-data-shapefile-units-from-recreation-model/2521/4 "2022-03-02T14:18:11Z")

</div>

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.