Mastering 3D Visualization with Python: Animation and Advanced Features

1. Animation and Advanced Python Features

2. Phase I Python is the best thing ever!

3. 3D Visualization: Positions particle configuration: “pos”

4. 3D Visualization: Scalars particle configuration: “pos” velocity configuration: “vel” velocity squared: “v2” “v2” on “pos” is like a scalar field.

5. 3D Visualization: Vectors particle configuration: “pos” velocity configuration: “vel” “vel” on “pos” is like a vector field.

6. 3D Visualization: Vectors + Scalars Composition is easy. Import from existing scripts. Protect main loop with if __name__ == '__main__':

7. 3D Animation: Positions all particle positions: “all_pos”

8. 3D Animation: Positions + Velocities all particle positions: “all_pos” all particle velocities: “all_vel”

9. Summary: 3D Visualization Snapshot Animation • Positions: • plot • scatter • Scalar field: • scatter with color • Vector field: • plot with quiver • Positions: • update Line3D artist • Scalar field: • update Path3DCollection artist • Vector field: • update Line3DCollection artist Don’t know how to update a particular artist? Use ArtistAnimation.

10. 2D Visualization: Scalars spin lattice: “lat”

11. Output to html spin lattice: “lat” We can upload your html videos to the course website under projects.

12. Phase II Python is the worst thing ever!

13. For loop: O(1000) times slower than fortran Autocorrelation function pyanswer, fanswer 15.694856 15.694856 pytime, ftime, py/f 0.039499 2.8e-05 1411 formula translation (fortran) was built with scientific computation in mind. Python implementation fortran implementation

14. For loop: O(1000) times slower than fortran Distance table N=256 pytime, ftime, py/f 0.199822 0.000391 511 Python implementation fortran implementation

15. Vectorize: O(10) times slower than fortran Autocorrelation function vecanswer, fanswer 15.694856 15.694856 vectime, ftime, vec/f 0.000368 2.6e-05 14 Vectorized Python implementation fortran implementation

16. Vectorize: O(10) times slower than fortran Distance table N=256 vectime, ftime, vec/f 0.004272 0.000387 11 Vectorized Python implementation fortran implementation

17. Vectorize: Q/ Where to put conditionals? Distance table Vectorized Python implementation fortran implementation

18. Phase III Embracing Python

19. Vectorize: Q/ Where to put conditionals? A/ Boolean Selector. Distance table Vectorized Python implementation fortran implementation

20. Boolean selector: array of True and False of same shape

21. List Comprehension: flatten for loop, remove append Cubic Posnatom=64,000 pytime, itertime, py/iter 0.0173 0.0062 2.8 For implementation Iter implementation Main loop

22. np.meshgrid: np.array instead of lists Cubic Posnatom=64,000 pytime, mtime, py/m 0.0173 0.0011 15.7 For implementation meshgrid implementation Main loop

23. np.einsum: general tensor contraction (Einstein notation) pos \cdotkvecs pytime, eintime, py/ein 0.005 0.0006 83

24. f2py: compile fortran modules and import md.f90 makefile Type “make” to create the library file md.so Then import into Python

25. Conclusion: Pick the right tool for the job • Python is a versatile language. It excels at: • Text parsing • Graphics • Composing available packaged functions • Interface with other codes • Python is not efficient at: • For loops • Memory management • Multiple Instruction Multiple Data (MIMD) Parallelization • In high-performance computing (HPC), consider Python as an interface • between backend performance code and the user. C, Fortran Data Structures Python Next step: analysis, graphics, ML, etc. f2py