1

I have a small project where I'd like to investigate the relationship between number of events (many thousand x-y points) in a city and the rent of each neighbourhood (around 50 x-y centroid points of neighbourhood polygons, each with a corresponding rent value).

I have these two data sets in two csvs (one with all the events and their lat/long, the other with all the neighbourhood centroids and their rent value).

I was hoping I would be able to click some kind of magic button in QGIS to do a geographically weighted regression (I already did a basic non-GW regression and plot in R and there was a nice relationship) in order to account for the proximity of the actual points for events/neighbourhoods... but I can't see an option anywhere in the menu for GWR.

Therefore,

1) Is there a way to do this easily in QGIS? Perhaps a plugin of some sort?

2) If not, do I have to use R and what is the plugin for this?

I've never done GWR before.

closed as too broad by Midavalo Feb 23 '17 at 21:22

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • You can use R scripts in the Processing Toolbox – gene Feb 23 '17 at 16:44
  • Ah yes thanks - is there one you suggest for basic GWR? – the_t_test_1 Feb 23 '17 at 16:49
  • You need a solution in R first and after, create the script ( example in Integrating QGIS and R: A stratified sampling example and QGIS-Processing rscripts) – gene Feb 23 '17 at 17:56
  • 1
    As per the Tour there should be only one question asked per question do it sounds like you should focus this on R and if you get stuck on QGIS later then ask about that then in a separate question. – PolyGeo Feb 23 '17 at 20:07
  • 1
    Please note that your neighborhoods represent areas, which may vary considerably in size, and cannot be treated as a discrete point feature. This is statistically invalid because a point does not represent the polygon area and the inference derived from the model would thus be incorrect. – Jeffrey Evans Feb 23 '17 at 22:05
1

Have you tried the standalone GWR4 app? It was developed by Fotheringham et al. It seems that it is missing a permanent home on the web right now, but I found this link: https://gwrtools.github.io/gwr4-downloads.html

Edit: The application is outside of QGIS and R but will accept CSV format as input. The outputs are a .txt file with summary information and a .csv file with local estimates for all of your points.

We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed.

  • Welcome to GIS SE! As a new user please take the tour to learn about our focused Q&A format. Please edit your answer to include some information about what your suggestion does, and how to use it in order to answer the question. – Midavalo Feb 23 '17 at 21:34

Not the answer you're looking for? Browse other questions tagged or ask your own question.