图片解析应用
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

53 lines
2.0 KiB

  1. import math
  2. from sympy.core.symbol import symbols
  3. from sympy.functions.elementary.exponential import exp
  4. from sympy.codegen.rewriting import optimize
  5. from sympy.codegen.approximations import SumApprox, SeriesApprox
  6. def test_SumApprox_trivial():
  7. x = symbols('x')
  8. expr1 = 1 + x
  9. sum_approx = SumApprox(bounds={x: (-1e-20, 1e-20)}, reltol=1e-16)
  10. apx1 = optimize(expr1, [sum_approx])
  11. assert apx1 - 1 == 0
  12. def test_SumApprox_monotone_terms():
  13. x, y, z = symbols('x y z')
  14. expr1 = exp(z)*(x**2 + y**2 + 1)
  15. bnds1 = {x: (0, 1e-3), y: (100, 1000)}
  16. sum_approx_m2 = SumApprox(bounds=bnds1, reltol=1e-2)
  17. sum_approx_m5 = SumApprox(bounds=bnds1, reltol=1e-5)
  18. sum_approx_m11 = SumApprox(bounds=bnds1, reltol=1e-11)
  19. assert (optimize(expr1, [sum_approx_m2])/exp(z) - (y**2)).simplify() == 0
  20. assert (optimize(expr1, [sum_approx_m5])/exp(z) - (y**2 + 1)).simplify() == 0
  21. assert (optimize(expr1, [sum_approx_m11])/exp(z) - (y**2 + 1 + x**2)).simplify() == 0
  22. def test_SeriesApprox_trivial():
  23. x, z = symbols('x z')
  24. for factor in [1, exp(z)]:
  25. x = symbols('x')
  26. expr1 = exp(x)*factor
  27. bnds1 = {x: (-1, 1)}
  28. series_approx_50 = SeriesApprox(bounds=bnds1, reltol=0.50)
  29. series_approx_10 = SeriesApprox(bounds=bnds1, reltol=0.10)
  30. series_approx_05 = SeriesApprox(bounds=bnds1, reltol=0.05)
  31. c = (bnds1[x][1] + bnds1[x][0])/2 # 0.0
  32. f0 = math.exp(c) # 1.0
  33. ref_50 = f0 + x + x**2/2
  34. ref_10 = f0 + x + x**2/2 + x**3/6
  35. ref_05 = f0 + x + x**2/2 + x**3/6 + x**4/24
  36. res_50 = optimize(expr1, [series_approx_50])
  37. res_10 = optimize(expr1, [series_approx_10])
  38. res_05 = optimize(expr1, [series_approx_05])
  39. assert (res_50/factor - ref_50).simplify() == 0
  40. assert (res_10/factor - ref_10).simplify() == 0
  41. assert (res_05/factor - ref_05).simplify() == 0
  42. max_ord3 = SeriesApprox(bounds=bnds1, reltol=0.05, max_order=3)
  43. assert optimize(expr1, [max_ord3]) == expr1