Adds an optional end point distance threshold parameter to the
network analysis tools. Previously (and still, by default) endpoints
will ALWAYS be snapped to the nearest point in the network layer,
regardless of how far away from the network they actually are. This
can result in meaningless results, as the tools will happily snap
points to a road hundreds of kilometers away :)
Now, there's an optional end point distance threshold parameter
for these tools. The behaviour of the threshold depends on the
algorithm:
- For the “Service area (from layer)” tool an optional new output
was added for “non routable features”. This output will contain
any features which were deemed too far from the network. All
other features which are within tolerance distance to the
network will be stored in the standard output from the tool.
- For the “Service area (from point)” tool an error will be raised
if the point is too far from the network
- For the “Shortest path (point to point)” tool an error will
be raised if either the source or destination points are too far
from the network.
- For the “Shortest path (layer to point)” and “Shortest path (point
to layer)” tools:
- An error will be raised if the **point** is too far from the network.
- A new optional output was added for “non routable features”. This
output will contain any features which were deemed too far from the
network. All other features which are within tolerance distance to
the network will be stored in the standard output from the tool.
Sponsored by City of Canning
Use the processing context's ellipsoid instead of a hardcoded
WGS84 ellipsoid for distance calculations during network
analysis, so that the lengths used will exactly match other
measurement tools used on the same features in the same
project.
There's two motivations for this:
- the existing one was getting massive and took ages to run, which was
a pain when developing. Smaller batches allow just a subset of test to
be run which is much faster.
- There's a random segfault on test exit which occurs on Travis. Rather
then disabling these absolutely critical tests altogether, I'm using
this as a method of bisecting exactly which alg is causing this.
