In vispy library I have a list of points shown as markers, and I want to change the color of(or just get the index of) the point which is nearest to the clicked point.
I can get the pixels of the clicked point by event.pos, but i need its actual co-ordinate to compare it with others (or get pixel point of other markers to compare it with the event location).
I have this code to get the nearest point index.which takes input of an array and a point(clicked one)
def get_nearest_index(pos,p0):
count=0
col =[(1,1,1,0.5) for i in pos]
dist= math.inf
ind=0
for i in range(len(pos)):
d = (pos[i][0]-p0[0])**2+(pos[i][1]-p0[1])**2
if d<dist:
ind=i
dist=d
return ind
But the problem is i have to pass both of them in same co-ordinate system.
Printing out event.pos returns pixels like: [319 313] while my positions, in pos array are:
[[-0.23801816 0.55117583 -0.56644607]
[-0.91117247 -2.28957391 -1.3636486 ]
[-1.81229627 0.50565064 -0.06175591]
[-1.79744952 0.48388072 -0.00389405]
[ 0.33729051 -0.4087148 0.57522977]]
So I need to convert one of them to another. Transformation like
tf = view.scene.transform
p0 = tf.map(pixel_pt)
print(str(pixel_pt) + "--->"+str(p0))
prints out [285 140 0 1]--->[ 4.44178173e+04 -1.60156369e+04 0.00000000e+00 1.00000000e+00] which is nowhere near the points.
When transforming the pixels to your local co-ordinates, you are using transform.map, which according to the vispy tutorial, gives you map co-ordinates. What you need to use is the inverse map.
You could try doing this:
tf = view.scene.transform
point = tf.imap(event.pos)
print(str(event.pos) + "--->"+str(point))
Likewise, in case you need to transform specific sets of markers, this would be a better approach.
ct = markers.node_transform(markers.root_node)
point = ct.imap(event.pos)