I was getting this problem with a complex map so I have created a simplified map with two layers.

Accessible greenspace contains a couple of hundred small, mostly isolated, polygons with one (name) data field.

Black Country 32 Towns has 32 larger contiguous polygons with three data fields.

They have the same CRS and physically all of the greenspace layer overlaps the towns layer.

I want to create a new layer the as as accessible greenspace, but with the appropriate town data appended to the polygons. Then I can use groupstats to find out how much accessible greenspace is in each town.

Intersect creates a new layer with all four data fields as I would expect, but it is empty. It should contain all the accessible greenspace polygons, with any overlapping different towns split into two.

Clips doesn't work either.

a test with buffer works, so the whole geoprocessing module isn't broken.

This sort of operation worked two weeks ago, but doesn't now.

Lots of questions here report a similar failure to create an output layer, but there seem to be not definitive answers, just requests for more detail or the suggestion to use GRASS (I find GRASS impenetrable).

My question is "Are the QGIS Intersect and Clip Tools Broken?" and is this an intermittent/random issue?

  • Have you checked the data for invalid polygons?
    – jpmc26
    Commented Jan 6, 2016 at 18:44
  • Which QGIS version are you using?
    – ndawson
    Commented Jan 6, 2016 at 19:57
  • As the previous commentators have asked, you can check for invalid shapes using Vector > Geometry Tools > Check Geometry Validity. Also from this post, I experienced a problem with QGIS 2.10 which wasn't able to perform the Difference tool effectively compared to other QGIS versions. You could consider uploading your shapefiles to a public server so that others could test your data.
    – Joseph
    Commented Jan 7, 2016 at 11:17
  • Yes, I have one duplicate point, which I removed. 2-12.1 Lyon I will upload example shapefiles that work and don't work tomorrow. Commented Jan 7, 2016 at 17:33

1 Answer 1


The short answer is no they are not broken. I've just checked with my data and they work fine.

So longer answer - there may well be an issue with in order of decreasing likelihood:

  1. your data
  2. your workflow
  3. your understanding of how the process works
  4. your install of QGis
  5. a combination of two of the above
  6. a combination of three of the above
  7. an actual bug with QGis

While it seems mean that people (here and on the mailing lists etc.) always ask for more details or propose a different way of solving the problem, our problem is that debugging a problem based on a short textual description of a problem is almost impossible. Please take the Tour to learn how this site works and if possible read through "How to ask questions the smart way" or if you really feel unloved watch my talk from FOSS4G15 which attempts to explain why questions like yours are not a favourite of developers.

  • I think it may be the data, as the command works as expected with some shapefiles. Commented Jan 7, 2016 at 17:30
  • Seems I timed out... I will upload some test shapefiles tomorrow. Commented Jan 7, 2016 at 17:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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