what is wrong with this shapely error message

Question

I have a dictionary called Rsize which have number-List as key-value pair. The dictionary is like this

{10: [0.6621485767296484, 0.6610747762560114, 0.659607022086639, 0.6567761845867727, 0.6535392433801197, 0.6485977028504701, 0.6393024556394106, 0.6223866436257335, 0.5999232392636733, 0.5418403536642005, 0.4961461379219235, 0.4280278015788386, 0.35462315989740956, 0.2863017237662875, 0.2312185739351389, 0.18306363413831017], 12: [0.6638977494825118, 0.663295576452323, 0.662262804664348, 0.6610413916318628, 0.6590939627030634, 0.655212304186114, 0.6492141689834672, 0.6380632834031537, 0.6096663492242224, 0.5647498006858608, 0.4983281599318278, 0.3961350546063216, 0.32119092575707087, 0.2257230704567207, 0.1816695139119151, 0.14363448808684576], 14: [0.6649598494971014, 0.6644370245269158, 0.6638578972784479, 0.6630511299276417, 0.6615070373022596, 0.6596206155163766, 0.6560628158033714, 0.6487119276511941, 0.6343385358239866, 0.5792725000508062, 0.49799837531709923, 0.42482204326408324, 0.26633662071414366, 0.2028085235063155, 0.12411214668987203, 0.09336935548451253]}
[0.6621485767296484, 0.6610747762560114, 0.659607022086639, 0.6567761845867727, 0.6535392433801197, 0.6485977028504701, 0.6393024556394106, 0.6223866436257335, 0.5999232392636733, 0.5418403536642005, 0.4961461379219235, 0.4280278015788386, 0.35462315989740956, 0.2863017237662875, 0.2312185739351389, 0.18306363413831017]

The keys are 10,14,16. I have used each list for plotting and want to find out their pairwise intersection points. I have written the following script for that and used shapely intersection function for the intersection points detection.

from shapely.geometry import LineString
Rsize={10: [0.6621485767296484, 0.6610747762560114, 0.659607022086639, 0.6567761845867727, 0.6535392433801197, 0.6485977028504701, 0.6393024556394106, 0.6223866436257335, 0.5999232392636733, 0.5418403536642005, 0.4961461379219235, 0.4280278015788386, 0.35462315989740956, 0.2863017237662875, 0.2312185739351389, 0.18306363413831017], 12: [0.6638977494825118, 0.663295576452323, 0.662262804664348, 0.6610413916318628, 0.6590939627030634, 0.655212304186114, 0.6492141689834672, 0.6380632834031537, 0.6096663492242224, 0.5647498006858608, 0.4983281599318278, 0.3961350546063216, 0.32119092575707087, 0.2257230704567207, 0.1816695139119151, 0.14363448808684576], 14: [0.6649598494971014, 0.6644370245269158, 0.6638578972784479, 0.6630511299276417, 0.6615070373022596, 0.6596206155163766, 0.6560628158033714, 0.6487119276511941, 0.6343385358239866, 0.5792725000508062, 0.49799837531709923, 0.42482204326408324, 0.26633662071414366, 0.2028085235063155, 0.12411214668987203, 0.09336935548451253]}
[0.6621485767296484, 0.6610747762560114, 0.659607022086639, 0.6567761845867727, 0.6535392433801197, 0.6485977028504701, 0.6393024556394106, 0.6223866436257335, 0.5999232392636733, 0.5418403536642005, 0.4961461379219235, 0.4280278015788386, 0.35462315989740956, 0.2863017237662875, 0.2312185739351389, 0.18306363413831017]

listkT = np.arange(4.0,4.8,0.05)

print(Rsize[10])

plt.figure(figsize=(18, 10))
plt.title ('Binder cumulant for critical point')

plt.plot(listkT, Rsize[10], '-',label='Lattice sie 10')
plt.plot(listkT, Rsize[12], '-',label='Lattice sie 12')
plt.plot(listkT, Rsize[14], '-',label='Lattice sie 14')
plt.legend() 
plt.show()

curve_10=LineString(np.column_stack((listkT, Rsize[10])))
curve_12=LineString(np.column_stack((listkT, Rsize[12])))
curve_14=LineString(np.column_stack((listkT, Rsize[14])))

intersection12 = curve_10.intersection(curve_12)
intersection14 = curve_10.intersection(curve_14)

plt.plot(*LineString(intersection12).xy, 'o')
plt.plot(*LineString(intersection14).xy, 'o')

x12, y = LineString(intersection12).xy
x14, y = LineString(intersection14).xy
print(np.intersect1d(x12, x14))
print(x12,x14)

But shapely throws an AssertionError.

  File "C:\Users\Endeavour\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 827, in runfile
    execfile(filename, namespace)

  File "C:\Users\Endeavour\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 110, in execfile
    exec(compile(f.read(), filename, 'exec'), namespace)
  File "E:/Project/Codes/3D.py", line 118, in <module>
    plt.plot(*LineString(intersection12).xy, 'o')

  File "C:\Users\Endeavour\Anaconda3\lib\site-packages\shapely\geometry\linestring.py", line 48, in __init__
    self._set_coords(coordinates)

  File "C:\Users\Endeavour\Anaconda3\lib\site-packages\shapely\geometry\linestring.py", line 97, in _set_coords
    ret = geos_linestring_from_py(coordinates)

  File "shapely/speedups/_speedups.pyx", line 87, in shapely.speedups._speedups.geos_linestring_from_py

AssertionError

The plots are drawn correctly by matplotlib though. I am using shapely for first time with no prior experience in it. Any help will be much appreciated. Thank you. Note: The final goal is to get the intersection of 3 curves. If no intersection found the point where they come closest is good enough. Any suggestion or library function to find that will be of great help.

Thank you in advance.

Answer 1

Following the assertion error, I checked shapely/speedups/_speedups.pyx, line 87. geos_linestring_from_py function expects you to either pass a LineString or a LinearRing. When I print your intersection12 and intersection14 I get:

POINT (4.503201814825258 0.4917840919384173)
POINT (4.51830999373466 0.4712012116887737)

So you are passing a Point instance to create a LineString, which creates an AssertionError.

Aside from the error you have, your approach is also wrong because it assumes that (1) there will be multiple intersections between two curves, and (2) there will be one absolute point where three curves intersect. If you zoom into your plot, you can see that neither is the case.

The red circle corresponds to your intersection12 and the purple one is intersection14. If you are looking for an approximate solution, maybe finding the mean of these points can help in this situation, but for more complex curves with multiple intersections per pair, it is also not recommended.