def fbenchmark(f, var=[Symbol('x')]):
"""
Do some benchmarks with f using clambdify, lambdify and psyco.
"""
global cf, pf, psyf
start = time()
cf = clambdify(var, f)
print('compile time (including sympy overhead): %f s' % (
time() - start))
pf = lambdify(var, f, 'math')
psyf = None
psyco = import_module('psyco')
if psyco:
psyf = lambdify(var, f, 'math')
psyco.bind(psyf)
code = '''for x in (i/1000. for i in range(1000)):
f(%s)''' % ('x,'*len(var)).rstrip(',')
t1 = Timer(code, 'from __main__ import cf as f')
t2 = Timer(code, 'from __main__ import pf as f')
if psyf:
t3 = Timer(code, 'from __main__ import psyf as f')
else:
t3 = None
print('for x = (0, 1, 2, ..., 999)/1000')
print('20 times in 3 runs')
print('compiled: %.4f %.4f %.4f' % tuple(t1.repeat(3, 20)))
print('Python lambda: %.4f %.4f %.4f' % tuple(t2.repeat(3, 20)))
if t3:
print('Psyco lambda: %.4f %.4f %.4f' % tuple(t3.repeat(3, 20)))
def mplot3d(f, var1, var2, show=True):
"""
Plot a 3d function using matplotlib/Tk.
"""
import warnings
warnings.filterwarnings("ignore", "Could not match \S")
p = import_module('pylab')
# Try newer version first
p3 = import_module('mpl_toolkits.mplot3d',
__import__kwargs={'fromlist':['something']}) or import_module('matplotlib.axes3d')
if not p or not p3:
sys.exit("Matplotlib is required to use mplot3d.")
x, y, z = sample(f, var1, var2)
fig = p.figure()
ax = p3.Axes3D(fig)
#ax.plot_surface(x,y,z) #seems to be a bug in matplotlib
ax.plot_wireframe(x,y,z)
ax.set_xlabel('X')
ax.set_ylabel('Y')
ax.set_zlabel('Z')
if show:
p.show()
def test_entropy():
up = JzKet(S(1)/2, S(1)/2)
down = JzKet(S(1)/2, -S(1)/2)
d = Density((up, 0.5), (down, 0.5))
# test for density object
ent = entropy(d)
assert entropy(d) == 0.5*log(2)
assert d.entropy() == 0.5*log(2)
np = import_module('numpy', min_module_version='1.4.0')
if np:
#do this test only if 'numpy' is available on test machine
np_mat = represent(d, format='numpy')
ent = entropy(np_mat)
assert isinstance(np_mat, np.matrixlib.defmatrix.matrix)
assert ent.real == 0.69314718055994529
assert ent.imag == 0
scipy = import_module('scipy', __import__kwargs={'fromlist': ['sparse']})
if scipy and np:
#do this test only if numpy and scipy are available
mat = represent(d, format="scipy.sparse")
assert isinstance(mat, scipy_sparse_matrix)
assert ent.real == 0.69314718055994529
assert ent.imag == 0
def test_is_scalar_sparse_matrix():
np = import_module('numpy')
if not np:
skip("numpy not installed.")
scipy = import_module('scipy', __import__kwargs={'fromlist': ['sparse']})
if not scipy:
skip("scipy not installed.")
numqubits = 2
id_only = False
id_gate = (IdentityGate(1),)
assert is_scalar_sparse_matrix(id_gate, numqubits, id_only) is True
x0 = X(0)
xx_circuit = (x0, x0)
assert is_scalar_sparse_matrix(xx_circuit, numqubits, id_only) is True
x1 = X(1)
y1 = Y(1)
xy_circuit = (x1, y1)
assert is_scalar_sparse_matrix(xy_circuit, numqubits, id_only) is False
z1 = Z(1)
xyz_circuit = (x1, y1, z1)
assert is_scalar_sparse_matrix(xyz_circuit, numqubits, id_only) is True
cnot = CNOT(1, 0)
cnot_circuit = (cnot, cnot)
assert is_scalar_sparse_matrix(cnot_circuit, numqubits, id_only) is True
h = H(0)
hh_circuit = (h, h)
assert is_scalar_sparse_matrix(hh_circuit, numqubits, id_only) is True
# NOTE:
# The elements of the sparse matrix for the following circuit
# is actually 1.0000000000000002+0.0j.
h1 = H(1)
xhzh_circuit = (x1, h1, z1, h1)
assert is_scalar_sparse_matrix(xhzh_circuit, numqubits, id_only) is True
id_only = True
assert is_scalar_sparse_matrix(xhzh_circuit, numqubits, id_only) is True
assert is_scalar_sparse_matrix(xyz_circuit, numqubits, id_only) is False
assert is_scalar_sparse_matrix(cnot_circuit, numqubits, id_only) is True
assert is_scalar_sparse_matrix(hh_circuit, numqubits, id_only) is True
def test_valued_tensor_self_contraction():
numpy = import_module("numpy")
if numpy is None:
skip("numpy not installed.")
assert AB(i0, -i0) == 4
assert BA(i0, -i0) == 2
def parse_latex(s):
r"""Converts the string ``s`` to a SymPy ``Expr``
Parameters
==========
s : str
The LaTeX string to parse. In Python source containing LaTeX,
*raw strings* (denoted with ``r"``, like this one) are preferred,
as LaTeX makes liberal use of the ``\`` character, which would
trigger escaping in normal Python strings.
Examples
========
>>> from sympy.parsing.latex import parse_latex # doctest: +SKIP
>>> expr = parse_latex(r"\frac {1 + \sqrt {\a}} {\b}") # doctest: +SKIP
>>> expr # doctest: +SKIP
(sqrt(a) + 1)/b
>>> expr.evalf(4, subs=dict(a=5, b=2)) # doctest: +SKIP
1.618
"""
_latex = import_module(
'sympy.parsing.latex._parse_latex_antlr',
__import__kwargs={'fromlist': ['X']})
if _latex is not None:
return _latex.parse_latex(s)
def test_valued_non_diagonal_metric():
numpy = import_module("numpy")
if numpy is None:
skip("numpy not installed.")
mmatrix = Matrix(ndm_matrix)
assert NA(n0)*NA(-n0) == (NA(n0).get_matrix().T * mmatrix * NA(n0).get_matrix())[0, 0]
def test_valued_tensor_applyfunc():
numpy = import_module("numpy")
if numpy is None:
skip("numpy not installed.")
aA = A(i0).applyfunc(lambda x: x**2)
aB = B(i0).applyfunc(lambda x: x**3)
aB2 = B(-i0).applyfunc(lambda x: x**3)
for i in range(4):
assert aA[i] == A(i0)[i]**2
assert aB[i] == B(i1)[i]**3
assert aB*aB2 == -794
tA = A.applyfunc(lambda x: x + 33)
tB = B.applyfunc(lambda x: x + 33)
tAB = AB.applyfunc(lambda x: x + 33)
assert (tA(i0)*tA(-i0)).expand() == ((E + 33)**2 - (px + 33)**2 - (py + 33)**2 - (pz + 33)**2).expand()
assert tB(i0).get_matrix() == Matrix([33, 34, 35, 36])
assert tAB(i0, i1).get_matrix() == Matrix([
[34, 33, 33, 33],
[33, 32, 33, 33],
[33, 33, 32, 33],
[33, 33, 33, 32],
])
def _get_meshes_grid(self):
"""Generates the mesh for generating a contour.
In the case of equality, ``contour`` function of matplotlib can
be used. In other cases, matplotlib's ``contourf`` is used.
"""
equal = False
if isinstance(self.expr, Equality):
expr = self.expr.lhs - self.expr.rhs
equal = True
elif isinstance(self.expr, (GreaterThan, StrictGreaterThan)):
expr = self.expr.lhs - self.expr.rhs
elif isinstance(self.expr, (LessThan, StrictLessThan)):
expr = self.expr.rhs - self.expr.lhs
else:
raise NotImplementedError("The expression is not supported for " "plotting in uniform meshed plot.")
np = import_module("numpy")
xarray = np.linspace(self.start_x, self.end_x, self.nb_of_points)
yarray = np.linspace(self.start_y, self.end_y, self.nb_of_points)
x_grid, y_grid = np.meshgrid(xarray, yarray)
func = vectorized_lambdify((self.var_x, self.var_y), expr)
z_grid = func(x_grid, y_grid)
z_grid[np.ma.where(z_grid < 0)] = -1
z_grid[np.ma.where(z_grid > 0)] = 1
if equal:
return xarray, yarray, z_grid, "contour"
else:
return xarray, yarray, z_grid, "contourf"
def init_ipython_session(argv=[], auto_symbols=False, auto_int_to_Integer=False):
"""Construct new IPython session. """
import IPython
if IPython.__version__ >= '0.11':
# use an app to parse the command line, and init config
# IPython 1.0 deprecates the frontend module, so we import directly
# from the terminal module to prevent a deprecation message from being
# shown.
if IPython.__version__ >= '1.0':
from IPython.terminal import ipapp
else:
from IPython.frontend.terminal import ipapp
app = ipapp.TerminalIPythonApp()
# don't draw IPython banner during initialization:
app.display_banner = False
app.initialize(argv)
if auto_symbols:
readline = import_module("readline")
if readline:
enable_automatic_symbols(app)
if auto_int_to_Integer:
enable_automatic_int_sympification(app)
return app.shell
else:
from IPython.Shell import make_IPython
return make_IPython(argv)
def test_evalonarray_numpy():
numpy = import_module('numpy')
a = numpy.arange(10, dtype=float)
evalonarray('lambda x: x + 1', a)
for i, j in enumerate(a):
if float(i + 1) != j:
raise ValueError("Values should be equal")
def parse_latex(sympy):
antlr4 = import_module('antlr4', warn_not_installed=True)
if None in [antlr4, MathErrorListener]:
raise ImportError("LaTeX parsing requires the antlr4 python package,"
" provided by pip (antlr4-python2-runtime or"
" antlr4-python3-runtime) or"
" conda (antlr-python-runtime)")
matherror = MathErrorListener(sympy)
stream = antlr4.InputStream(sympy)
lex = LaTeXLexer(stream)
lex.removeErrorListeners()
lex.addErrorListener(matherror)
tokens = antlr4.CommonTokenStream(lex)
parser = LaTeXParser(tokens)
# remove default console error listener
parser.removeErrorListeners()
parser.addErrorListener(matherror)
relation = parser.math().relation()
expr = convert_relation(relation)
return expr
def test_issue_15265():
from sympy.core.sympify import sympify
from sympy.core.singleton import S
matplotlib = import_module('matplotlib', min_module_version='1.1.0', catch=(RuntimeError,))
if not matplotlib:
skip("Matplotlib not the default backend")
x = Symbol('x')
eqn = sin(x)
p = plot(eqn, xlim=(-S.Pi, S.Pi), ylim=(-1, 1))
p._backend.close()
p = plot(eqn, xlim=(-1, 1), ylim=(-S.Pi, S.Pi))
p._backend.close()
p = plot(eqn, xlim=(-1, 1), ylim=(sympify('-3.14'), sympify('3.14')))
p._backend.close()
p = plot(eqn, xlim=(sympify('-3.14'), sympify('3.14')), ylim=(-1, 1))
p._backend.close()
raises(ValueError,
lambda: plot(eqn, xlim=(-S.ImaginaryUnit, 1), ylim=(-1, 1)))
raises(ValueError,
lambda: plot(eqn, xlim=(-1, 1), ylim=(-1, S.ImaginaryUnit)))
raises(ValueError,
lambda: plot(eqn, xlim=(-S.Infinity, 1), ylim=(-1, 1)))
raises(ValueError,
lambda: plot(eqn, xlim=(-1, 1), ylim=(-1, S.Infinity)))
def test_matplotlib():
matplotlib = import_module('matplotlib', min_module_version='1.1.0', catch=(RuntimeError,))
if matplotlib:
plot_implicit_tests('test')
test_line_color()
else:
skip("Matplotlib not the default backend")
def mplot2d(f, var, show=True):
"""
Plot a 2d function using matplotlib/Tk.
"""
import warnings
warnings.filterwarnings("ignore", "Could not match \S")
p = import_module('pylab')
if not p:
sys.exit("Matplotlib is required to use mplot2d.")
if not is_sequence(f):
f = [f,]
for f_i in f:
x, y = sample2d(f_i, var)
p.plot(x, y,'black')
p.draw()
p.ylabel("Transverse beam cordinate")
p.xlabel("Propagation coordinate")
p.grid(True)
if show:
p.show()
def cos(x):
"""Evaluates the cos of an interval"""
np = import_module('numpy')
if isinstance(x, (int, float)):
return interval(np.sin(x))
elif isinstance(x, interval):
if not (np.isfinite(x.start) and np.isfinite(x.end)):
return interval(-1, 1, is_valid=x.is_valid)
na, __ = divmod(x.start, np.pi / 2.0)
nb, __ = divmod(x.end, np.pi / 2.0)
start = min(np.cos(x.start), np.cos(x.end))
end = max(np.cos(x.start), np.cos(x.end))
if nb - na > 4:
#differ more than 2*pi
return interval(-1, 1, is_valid=x.is_valid)
elif na == nb:
#in the same quadarant
return interval(start, end, is_valid=x.is_valid)
else:
if (na) // 4 != (nb) // 4:
#cos has max
end = 1
if (na - 2) // 4 != (nb - 2) // 4:
#cos has min
start = -1
return interval(start, end, is_valid=x.is_valid)
else:
raise NotImplementedError
def sin(x):
"""evaluates the sine of an interval"""
np = import_module('numpy')
if isinstance(x, (int, float)):
return interval(np.sin(x))
elif isinstance(x, interval):
if not x.is_valid:
return interval(-1, 1, is_valid=x.is_valid)
na, __ = divmod(x.start, np.pi / 2.0)
nb, __ = divmod(x.end, np.pi / 2.0)
start = min(np.sin(x.start), np.sin(x.end))
end = max(np.sin(x.start), np.sin(x.end))
if nb - na > 4:
return interval(-1, 1, is_valid=x.is_valid)
elif na == nb:
return interval(start, end, is_valid=x.is_valid)
else:
if (na - 1) // 4 != (nb - 1) // 4:
#sin has max
end = 1
if (na - 3) // 4 != (nb - 3) // 4:
#sin has min
start = -1
return interval(start, end)
else:
raise NotImplementedError
def test_no_stdlib_collections3():
'''make sure we get the right collections with no catch'''
import collections
matplotlib = import_module('matplotlib',
__import__kwargs={'fromlist': ['cm', 'collections']},
min_module_version='1.1.0')
if matplotlib:
assert collections != matplotlib.collections
def __call__(self, *args):
np = import_module('numpy')
np_old_err = np.seterr(invalid='raise')
try:
temp_args = (np.array(a, dtype=np.complex) for a in args)
results = self.vector_func(*temp_args)
results = np.ma.masked_where(
np.abs(results.imag) > 1e-7 * np.abs(results),
results.real, copy=False)
except Exception as e:
#DEBUG: print 'Error', type(e), e
if ((isinstance(e, TypeError)
and 'unhashable type: \'numpy.ndarray\'' in str(e))
or
(isinstance(e, ValueError)
and ('Invalid limits given:' in str(e)
or 'negative dimensions are not allowed' in str(e) # XXX
or 'sequence too large; must be smaller than 32' in str(e)))): # XXX
# Almost all functions were translated to numpy, but some were
# left as sympy functions. They received an ndarray as an
# argument and failed.
# sin(ndarray(...)) raises "unhashable type"
# Integral(x, (x, 0, ndarray(...))) raises "Invalid limits"
# other ugly exceptions that are not well understood (marked with XXX)
# TODO: Cleanup the ugly special cases marked with xxx above.
# Solution: use cmath and vectorize the final lambda.
self.lambda_func = experimental_lambdify(
self.args, self.expr, use_python_cmath=True)
self.vector_func = np.vectorize(
self.lambda_func, otypes=[np.complex])
results = self.vector_func(*args)
results = np.ma.masked_where(
np.abs(results.imag) > 1e-7 * np.abs(results),
results.real, copy=False)
else:
# Complete failure. One last try with no translations, only
# wrapping in complex((...).evalf()) and returning the real
# part.
if self.failure:
raise e
else:
self.failure = True
self.lambda_func = experimental_lambdify(
self.args, self.expr, use_evalf=True,
complex_wrap_evalf=True)
self.vector_func = np.vectorize(
self.lambda_func, otypes=[np.complex])
results = self.vector_func(*args)
results = np.ma.masked_where(
np.abs(results.imag) > 1e-7 * np.abs(results),
results.real, copy=False)
warnings.warn('The evaluation of the expression is'
' problematic. We are trying a failback method'
' that may still work. Please report this as a bug.')
finally:
np.seterr(**np_old_err)
return results
def test_valued_tensor_pow():
numpy = import_module("numpy")
if numpy is None:
skip("numpy not installed.")
assert C**2 == -E**2 + px**2 + py**2 + pz**2
assert C**1 == sqrt(-E**2 + px**2 + py**2 + pz**2)
assert C(mu0)**2 == C**2
assert C(mu0)**1 == C**1
def test_bfs_identity_search_xfail():
scipy = import_module('scipy', __import__kwargs={'fromlist': ['sparse']})
if scipy:
skip("scipy installed.")
s = PhaseGate(0)
t = TGate(0)
gate_list = [Dagger(s), t]
id_set = {GateIdentity(Dagger(s), t, t)}
assert bfs_identity_search(gate_list, 1, max_depth=3) == id_set
def test_matrix_numpy():
from sympy.external import import_module
numpy = import_module('numpy')
if not numpy:
return
l = [[1, 2], [3, 4], [5, 6]]
a = numpy.matrix(l)
assert matrix(l) == matrix(a)
def exp(x):
"""evaluates the exponential of an interval"""
np = import_module('numpy')
if isinstance(x, (int, float)):
return interval(np.exp(x), np.exp(x))
elif isinstance(x, interval):
return interval(np.exp(x.start), np.exp(x.end), is_valid=x.is_valid)
else:
raise NotImplementedError
def tanh(x):
"""Evaluates the hyperbolic tan of an interval"""
np = import_module('numpy')
if isinstance(x, (int, float)):
return interval(np.tanh(x), np.tanh(x))
elif isinstance(x, interval):
return interval(np.tanh(x.start), np.tanh(x.end), is_valid=x.is_valid)
else:
raise NotImplementedError
def test_matplotlib():
matplotlib = import_module('matplotlib', min_module_version='1.1.0', catch=(RuntimeError,))
if matplotlib:
try:
plot_and_save('test')
finally:
# clean up
TmpFileManager.cleanup()
else:
skip("Matplotlib not the default backend")
def asinh(x):
"""Evaluates the inverse hyperbolic sine of an interval"""
np = import_module('numpy')
if isinstance(x, (int, float)):
return interval(np.arcsinh(x))
elif isinstance(x, interval):
start = np.arcsinh(x.start)
end = np.arcsinh(x.end)
return interval(start, end, is_valid=x.is_valid)
else:
return NotImplementedError
def atan(x):
"""evaluates the tan inverse of an interval"""
np = import_module('numpy')
if isinstance(x, (int, float)):
return interval(np.arctan(x))
elif isinstance(x, interval):
start = np.arctan(x.start)
end = np.arctan(x.end)
return interval(start, end, is_valid=x.is_valid)
else:
raise NotImplementedError
def test_issue_11463():
numpy = import_module('numpy')
if not numpy:
skip("numpy not installed.")
x = Symbol('x')
f = lambdify(x, real_root((log(x/(x-2))), 3), 'numpy')
# numpy.select evaluates all options before considering conditions,
# so it raises a warning about root of negative number which does
# not affect the outcome. This warning is suppressed here
with ignore_warnings(RuntimeWarning):
assert f(numpy.array(-1)) < -1
def test_no_stdlib_collections():
'''
make sure we get the right collections when it is not part of a
larger list
'''
import collections
matplotlib = import_module('matplotlib',
__import__kwargs={'fromlist': ['cm', 'collections']},
min_module_version='1.1.0', catch=(RuntimeError,))
if matplotlib:
assert collections != matplotlib.collections
def test_valued_tensor_covariant_contravariant_elements():
numpy = import_module("numpy")
if numpy is None:
skip("numpy not installed.")
assert A(-i0)[0] == A(i0)[0]
assert A(-i0)[1] == -A(i0)[1]
assert AB(i0, i1)[1, 1] == -1
assert AB(i0, -i1)[1, 1] == 1
assert AB(-i0, -i1)[1, 1] == -1
assert AB(-i0, i1)[1, 1] == 1
from __future__ import print_function, division
from sympy import Mul
from sympy.core.compatibility import u
from sympy.external import import_module
from sympy.physics.quantum.gate import Gate, OneQubitGate, CGate, CGateS
__all__ = [
'CircuitPlot',
'circuit_plot',
'labeller',
'Mz',
'Mx',
]
np = import_module('numpy')
matplotlib = import_module(
'matplotlib',
__import__kwargs={'fromlist': ['pyplot']},
catch=(RuntimeError,
)) # This is raised in environments that have no display.
if not np or not matplotlib:
class CircuitPlot(object):
def __init__(*args, **kwargs):
raise ImportError('numpy or matplotlib not available.')
def circuit_plot(*args, **kwargs):
raise ImportError('numpy or matplotlib not available.')
else:
def init_session(ipython=None, pretty_print=True, order=None,
use_unicode=None, use_latex=None, quiet=False, auto_symbols=False,
auto_int_to_Integer=False, argv=[]):
"""
Initialize an embedded IPython or Python session. The IPython session is
initiated with the --pylab option, without the numpy imports, so that
matplotlib plotting can be interactive.
Parameters
==========
pretty_print: boolean
If True, use pretty_print to stringify;
if False, use sstrrepr to stringify.
order: string or None
There are a few different settings for this parameter:
lex (default), which is lexographic order;
grlex, which is graded lexographic order;
grevlex, which is reversed graded lexographic order;
old, which is used for compatibility reasons and for long expressions;
None, which sets it to lex.
use_unicode: boolean or None
If True, use unicode characters;
if False, do not use unicode characters.
use_latex: boolean or None
If True, use latex rendering if IPython GUI's;
if False, do not use latex rendering.
quiet: boolean
If True, init_session will not print messages regarding its status;
if False, init_session will print messages regarding its status.
auto_symbols: boolean
If True, IPython will automatically create symbols for you.
If False, it will not.
The default is False.
auto_int_to_Integer: boolean
If True, IPython will automatically wrap int literals with Integer, so
that things like 1/2 give Rational(1, 2).
If False, it will not.
The default is False.
ipython: boolean or None
If True, printing will initialize for an IPython console;
if False, printing will initialize for a normal console;
The default is None, which automatically determines whether we are in
an ipython instance or not.
argv: list of arguments for IPython
See sympy.bin.isympy for options that can be used to initialize IPython.
See Also
========
sympy.interactive.printing.init_printing: for examples and the rest of the parameters.
Examples
========
>>> from sympy import init_session, Symbol, sin, sqrt
>>> sin(x) #doctest: +SKIP
NameError: name 'x' is not defined
>>> init_session() #doctest: +SKIP
>>> sin(x) #doctest: +SKIP
sin(x)
>>> sqrt(5) #doctest: +SKIP
___
\/ 5
>>> init_session(pretty_print=False) #doctest: +SKIP
>>> sqrt(5) #doctest: +SKIP
sqrt(5)
>>> y + x + y**2 + x**2 #doctest: +SKIP
x**2 + x + y**2 + y
>>> init_session(order='grlex') #doctest: +SKIP
>>> y + x + y**2 + x**2 #doctest: +SKIP
x**2 + y**2 + x + y
>>> init_session(order='grevlex') #doctest: +SKIP
>>> y * x**2 + x * y**2 #doctest: +SKIP
x**2*y + x*y**2
>>> init_session(order='old') #doctest: +SKIP
>>> x**2 + y**2 + x + y #doctest: +SKIP
x + y + x**2 + y**2
>>> theta = Symbol('theta') #doctest: +SKIP
>>> theta #doctest: +SKIP
theta
>>> init_session(use_unicode=True) #doctest: +SKIP
>>> theta # doctest: +SKIP
\u03b8
"""
import sys
in_ipython = False
if ipython is not False:
try:
import IPython
except ImportError:
if ipython is True:
raise RuntimeError("IPython is not available on this system")
ip = None
else:
if IPython.__version__ >= '0.11':
try:
ip = get_ipython()
except NameError:
ip = None
else:
ip = IPython.ipapi.get()
if ip:
ip = ip.IP
in_ipython = bool(ip)
if ipython is None:
ipython = in_ipython
if ipython is False:
ip = init_python_session()
mainloop = ip.interact
else:
if ip is None:
ip = init_ipython_session(argv=argv, auto_symbols=auto_symbols,
auto_int_to_Integer=auto_int_to_Integer)
if IPython.__version__ >= '0.11':
# runsource is gone, use run_cell instead, which doesn't
# take a symbol arg. The second arg is `store_history`,
# and False means don't add the line to IPython's history.
ip.runsource = lambda src, symbol='exec': ip.run_cell(src, False)
#Enable interactive plotting using pylab.
try:
ip.enable_pylab(import_all=False)
except Exception:
# Causes an import error if matplotlib is not installed.
# Causes other errors (depending on the backend) if there
# is no display, or if there is some problem in the
# backend, so we have a bare "except Exception" here
pass
if not in_ipython:
mainloop = ip.mainloop
readline = import_module("readline")
if auto_symbols and (not ipython or IPython.__version__ < '0.11' or not readline):
raise RuntimeError("automatic construction of symbols is possible only in IPython 0.11 or above with readline support")
if auto_int_to_Integer and (not ipython or IPython.__version__ < '0.11'):
raise RuntimeError("automatic int to Integer transformation is possible only in IPython 0.11 or above")
_preexec_source = preexec_source
ip.runsource(_preexec_source, symbol='exec')
init_printing(pretty_print=pretty_print, order=order,
use_unicode=use_unicode, use_latex=use_latex, ip=ip)
message = _make_message(ipython, quiet, _preexec_source)
if not in_ipython:
mainloop(message)
sys.exit('Exiting ...')
else:
ip.write(message)
import atexit
atexit.register(lambda ip: ip.write("Exiting ...\n"), ip)
from sympy.parsing.sym_expr import SymPyExpression
from sympy.testing.pytest import raises
from sympy.external import import_module
cin = import_module("clang.cindex", import_kwargs={"fromlist": ["cindex"]})
if cin:
from sympy.codegen.ast import (
Variable,
IntBaseType,
FloatBaseType,
String,
Return,
FunctionDefinition,
Integer,
Float,
Declaration,
CodeBlock,
FunctionPrototype,
FunctionCall,
NoneToken,
)
from sympy import Symbol
import os
def test_variable():
c_src1 = "int a;" + "\n" + "int b;" + "\n"
c_src2 = "float a;" + "\n" + "float b;" + "\n"
c_src3 = "int a;" + "\n" + "float b;" + "\n" + "int c;"
c_src4 = "int x = 1, y = 6.78;" + "\n" + "float p = 2, q = 9.67;"
besselk, S, beta, MatrixSymbol, chebyshevt, chebyshevu,
legendre, hermite, laguerre, gegenbauer, assoc_legendre,
assoc_laguerre, jacobi, fresnelc, fresnels)
from sympy.printing.lambdarepr import LambdaPrinter
from sympy.printing.pycode import NumPyPrinter
from sympy.utilities.lambdify import implemented_function, lambdastr
from sympy.utilities.pytest import skip
from sympy.utilities.decorator import conserve_mpmath_dps
from sympy.external import import_module
from sympy.functions.special.gamma_functions import uppergamma, lowergamma
import sympy
MutableDenseMatrix = Matrix
numpy = import_module('numpy')
scipy = import_module('scipy')
numexpr = import_module('numexpr')
tensorflow = import_module('tensorflow')
if tensorflow:
# Hide Tensorflow warnings
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
w, x, y, z = symbols('w,x,y,z')
#================== Test different arguments =======================
def test_no_args():
from functools import singledispatch
from sympy import Dummy, lambdify, exp
from sympy.external import import_module
from sympy.stats import DiscreteDistributionHandmade
from sympy.stats.crv import SingleContinuousDistribution
from sympy.stats.crv_types import ChiSquaredDistribution, ExponentialDistribution, GammaDistribution, \
LogNormalDistribution, NormalDistribution, ParetoDistribution, UniformDistribution, BetaDistribution, \
StudentTDistribution, CauchyDistribution
from sympy.stats.drv_types import GeometricDistribution, LogarithmicDistribution, NegativeBinomialDistribution, \
PoissonDistribution, SkellamDistribution, YuleSimonDistribution, ZetaDistribution
from sympy.stats.frv import SingleFiniteDistribution
scipy = import_module("scipy")
if scipy is not None:
import scipy.stats
@singledispatch
def do_sample_scipy(dist, size, seed):
return None
# CRV
@do_sample_scipy.register(SingleContinuousDistribution)
def _(dist: SingleContinuousDistribution, size, seed):
# if we don't need to make a handmade pdf, we won't
import scipy.stats
from sympy.external import import_module
matchpy = import_module("matchpy")
from sympy.utilities.decorator import doctest_depends_on
from sympy.core import Integer, Float
import inspect, re
if matchpy:
from matchpy import (Operation, CommutativeOperation, AssociativeOperation,
ManyToOneReplacer, OneIdentityOperation, CustomConstraint)
from matchpy.expressions.functions import register_operation_iterator, register_operation_factory
from sympy import Pow, Add, Integral, Basic, Mul, S
from sympy.functions import (log, sin, cos, tan, cot, csc, sec, sqrt, erf,
exp, log, gamma, acosh, asinh, atanh, acoth, acsch, asech, cosh, sinh,
tanh, coth, sech, csch, atan, acsc, asin, acot, acos, asec, fresnels,
fresnelc, erfc, erfi)
Operation.register(Integral)
register_operation_iterator(Integral, lambda a: (a._args[0],) + a._args[1], lambda a: len((a._args[0],) + a._args[1]))
Operation.register(Pow)
OneIdentityOperation.register(Pow)
register_operation_iterator(Pow, lambda a: a._args, lambda a: len(a._args))
Operation.register(Add)
OneIdentityOperation.register(Add)
CommutativeOperation.register(Add)
AssociativeOperation.register(Add)
register_operation_iterator(Add, lambda a: a._args, lambda a: len(a._args))
Operation.register(Mul)
OneIdentityOperation.register(Mul)
exp,
floor,
imax,
imin,
interval,
log,
log10,
Or,
sin,
sinh,
sqrt,
tan,
tanh,
)
np = import_module("numpy")
if not np:
disabled = True
# requires Numpy. Hence included in interval_functions
def test_interval_pow():
a = 2**interval(1, 2) == interval(2, 4)
assert a == (True, True)
a = interval(1, 2)**interval(1, 2) == interval(1, 4)
assert a == (True, True)
a = interval(-1, 1)**interval(0.5, 2)
assert a.is_valid is None
a = interval(-2, -1)**interval(1, 2)
assert a.is_valid is False
def benchmark():
"""
Run some benchmarks for clambdify and frange.
NumPy and Psyco are used as reference if available.
"""
from time import time
from timeit import Timer
def fbenchmark(f, var=[Symbol('x')]):
"""
Do some benchmarks with f using clambdify, lambdify and psyco.
"""
global cf, pf, psyf
start = time()
cf = clambdify(var, f)
print 'compile time (including sympy overhead): %f s' % (time() - start)
pf = lambdify(var, f, 'math')
psyf = None
psyco = import_module('psyco')
if psyco:
psyf = lambdify(var, f, 'math')
psyco.bind(psyf)
code = '''for x in (i/1000. for i in range(1000)):
f(%s)''' % ('x,'*len(var)).rstrip(',')
t1 = Timer(code, 'from __main__ import cf as f')
t2 = Timer(code, 'from __main__ import pf as f')
if psyf:
t3 = Timer(code, 'from __main__ import psyf as f')
else:
t3 = None
print 'for x = (0, 1, 2, ..., 999)/1000'
print '20 times in 3 runs'
print 'compiled: %.4f %.4f %.4f' % tuple(t1.repeat(3, 20))
print 'Python lambda: %.4f %.4f %.4f' % tuple(t2.repeat(3, 20))
if t3:
print 'Psyco lambda: %.4f %.4f %.4f' % tuple(t3.repeat(3, 20))
print 'big function:'
from sympy import diff, _exp, _sin, _cos, pi, lambdify
x = Symbol('x')
## f1 = diff(_exp(x)**2 - _sin(x)**pi, x) \
## * x**12-2*x**3+2*_exp(x**2)-3*x**7+4*_exp(123+x-x**5+2*x**4) \
## * ((x + pi)**5).expand()
f1 = 2*_exp(x**2) + x**12*(-pi*_sin(x)**((-1) + pi)*_cos(x) + 2*_exp(2*x)) \
+ 4*(10*pi**3*x**2 + 10*pi**2*x**3 + 5*pi*x**4 + 5*x*pi**4 + pi**5 \
+ x**5)*_exp(123 + x + 2*x**4 - x**5) - 2*x**3 - 3*x**7
fbenchmark(f1)
print
print 'simple function:'
y = Symbol('y')
f2 = sqrt(x*y)+x*5
fbenchmark(f2, [x,y])
times = 100000
fstr = '_exp(_sin(_exp(-x**2)) + sqrt(pi)*_cos(x**5/(x**3-x**2+pi*x)))'
print
print 'frange with f(x) ='
print fstr
print 'for x=1, ..., %i' % times
print 'in 3 runs including full compile time'
t4 = Timer("frange('lambda x: %s', 0, %i)" % (fstr, times),
'from __main__ import frange')
numpy = import_module('numpy')
print 'frange: %.4f %.4f %.4f' % tuple(t4.repeat(3, 1))
if numpy:
t5 = Timer('x = arange(%i); result = %s' % (times, fstr),
'from numpy import arange, sqrt, exp, sin, cos, exp, pi')
print 'numpy: %.4f %.4f %.4f' % tuple(t5.repeat(3, 1))
# Q: What about _imp_ functions?
# A: Those are taken care for by evalf. A special case treatment will work
# faster but it's not worth the code complexity.
# Q: Will ast fix all possible problems?
# A: No. You will always have to use some printer. Even srepr may not work in
# some cases. But if the printer does not work, that should be considered a
# bug.
# Q: Is there same way to fix all possible problems?
# A: Probably by constructing our strings ourself by traversing the (func,
# args) tree and creating the namespace at the same time. That actually sounds
# good.
from sympy.external import import_module
np = import_module('numpy')
import warnings
#TODO debuging output
class vectorized_lambdify(object):
""" Return a sufficiently smart, vectorized and lambdified function.
Returns only reals.
This function uses experimental_lambdify to created a lambdified
expression ready to be used with numpy. Many of the functions in sympy
are not implemented in numpy so in some cases we resort to python cmath or
even to evalf.
CodegenArrayPermuteDims,
CodegenArrayDiagonal)
from sympy.core.relational import Eq, Ne, Ge, Gt, Le, Lt
from sympy.external import import_module
from sympy.functions import \
Abs, Max, Min, ceiling, exp, floor, sign, sin, asin, sqrt, cos, \
acos, tan, atan, atan2, cosh, acosh, sinh, asinh, tanh, atanh, \
re, im, arg, erf, loggamma, log
from sympy.matrices import Matrix, MatrixBase, eye, randMatrix
from sympy.matrices.expressions import \
Determinant, HadamardProduct, Inverse, MatrixSymbol, Trace
from sympy.printing.tensorflow import TensorflowPrinter, tensorflow_code
from sympy.utilities.lambdify import lambdify
from sympy.utilities.pytest import skip
tf = tensorflow = import_module("tensorflow")
if tensorflow:
# Hide Tensorflow warnings
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
M = MatrixSymbol("M", 3, 3)
N = MatrixSymbol("N", 3, 3)
P = MatrixSymbol("P", 3, 3)
Q = MatrixSymbol("Q", 3, 3)
x, y, z, t = symbols("x y z t")
if tf is not None:
llo = [[j for j in range(i, i + 3)] for i in range(0, 9, 3)]
ad = RaisingOp('a')
a = LoweringOp('a')
k = SHOKet('k')
kz = SHOKet(0)
kf = SHOKet(1)
k3 = SHOKet(3)
b = SHOBra('b')
b3 = SHOBra(3)
H = Hamiltonian('H')
N = NumberOp('N')
omega = Symbol('omega')
m = Symbol('m')
ndim = Integer(4)
np = import_module('numpy')
scipy = import_module('scipy', __import__kwargs={'fromlist': ['sparse']})
ad_rep_sympy = represent(ad, basis=N, ndim=4, format='sympy')
a_rep = represent(a, basis=N, ndim=4, format='sympy')
N_rep = represent(N, basis=N, ndim=4, format='sympy')
H_rep = represent(H, basis=N, ndim=4, format='sympy')
k3_rep = represent(k3, basis=N, ndim=4, format='sympy')
b3_rep = represent(b3, basis=N, ndim=4, format='sympy')
def test_RaisingOp():
assert Dagger(ad) == a
assert Commutator(ad, a).doit() == Integer(-1)
assert Commutator(ad, N).doit() == Integer(-1) * ad
assert qapply(ad * k) == (sqrt(k.n + 1) * SHOKet(k.n + 1)).expand()
# Ported from latex2sympy by @augustt198
# https://github.com/augustt198/latex2sympy
# See license in LICENSE.txt
import sympy
from sympy.external import import_module
from sympy.printing.str import StrPrinter
from .errors import LaTeXParsingError
LaTeXParser = LaTeXLexer = MathErrorListener = None
try:
LaTeXParser = import_module('sympy.parsing.latex._antlr.latexparser',
__import__kwargs={
'fromlist': ['LaTeXParser']
}).LaTeXParser
LaTeXLexer = import_module('sympy.parsing.latex._antlr.latexlexer',
__import__kwargs={
'fromlist': ['LaTeXLexer']
}).LaTeXLexer
except Exception:
pass
ErrorListener = import_module('antlr4.error.ErrorListener',
warn_not_installed=True,
__import__kwargs={'fromlist': ['ErrorListener']})
if ErrorListener:
class MathErrorListener(ErrorListener.ErrorListener):
def sample_iter(expr,
condition=None,
size=(1, ),
library='scipy',
numsamples=S.Infinity,
**kwargs):
"""
Returns an iterator of realizations from the expression given a condition
Parameters
==========
expr: Expr
Random expression to be realized
condition: Expr, optional
A conditional expression
size : int, tuple
Represents size of each sample in numsamples
numsamples: integer, optional
Length of the iterator (defaults to infinity)
Examples
========
>>> from sympy.stats import Normal, sample_iter
>>> X = Normal('X', 0, 1)
>>> expr = X*X + 3
>>> iterator = sample_iter(expr, numsamples=3) # doctest: +SKIP
>>> list(iterator) # doctest: +SKIP
[12, 4, 7]
Returns
=======
sample_iter: iterator object
iterator object containing the sample/samples of given expr
See Also
========
sample
sampling_P
sampling_E
"""
if not import_module(library):
raise ValueError("Failed to import %s" % library)
if condition is not None:
ps = pspace(Tuple(expr, condition))
else:
ps = pspace(expr)
rvs = list(ps.values)
if library == 'pymc3':
# Currently unable to lambdify in pymc3
# TODO : Remove 'pymc3' when lambdify accepts 'pymc3' as module
fn = lambdify(rvs, expr, **kwargs)
else:
fn = lambdify(rvs, expr, modules=library, **kwargs)
if condition is not None:
given_fn = lambdify(rvs, condition, **kwargs)
def return_generator():
count = 0
while count < numsamples:
d = ps.sample(
size=size,
library=library) # a dictionary that maps RVs to values
args = [d[rv] for rv in rvs]
if condition is not None: # Check that these values satisfy the condition
gd = given_fn(*args)
if gd != True and gd != False:
raise ValueError(
"Conditions must not contain free symbols")
if not gd: # If the values don't satisfy then try again
continue
yield fn(*args)
count += 1
return return_generator()
import sympy
import tempfile
import os
from sympy import symbols, Eq, Mod
from sympy.external import import_module
from sympy.tensor import IndexedBase, Idx
from sympy.utilities.autowrap import autowrap, ufuncify, CodeWrapError
from sympy.utilities.pytest import skip
numpy = import_module('numpy', min_module_version='1.6.1')
Cython = import_module('Cython', min_module_version='0.15.1')
f2py = import_module('numpy.f2py', __import__kwargs={'fromlist': ['f2py']})
f2pyworks = False
if f2py:
try:
autowrap(symbols('x'), 'f95', 'f2py')
except (CodeWrapError, ImportError, OSError):
f2pyworks = False
else:
f2pyworks = True
a, b, c = symbols('a b c')
n, m, d = symbols('n m d', integer=True)
A, B, C = symbols('A B C', cls=IndexedBase)
i = Idx('i', m)
j = Idx('j', n)
k = Idx('k', d)
def has_module(module):
def test_evalonarray_numpy():
numpy = import_module('numpy')
a = numpy.arange(10, dtype=float)
evalonarray('lambda x: x + 1', a)
for i, j in enumerate(a):
assert float(i + 1) == j
# Ported from latex2sympy by @augustt198
# https://github.com/augustt198/latex2sympy
# See license in LICENSE.txt
import sympy
from sympy.external import import_module
from sympy.printing.str import StrPrinter
from .errors import LaTeXParsingError
LaTeXParser = LaTeXLexer = MathErrorListener = None
try:
LaTeXParser = import_module(
"sympy.parsing.latex._antlr.latexparser",
import_kwargs={
"fromlist": ["LaTeXParser"]
},
).LaTeXParser
LaTeXLexer = import_module(
"sympy.parsing.latex._antlr.latexlexer",
import_kwargs={
"fromlist": ["LaTeXLexer"]
},
).LaTeXLexer
except Exception:
pass
ErrorListener = import_module(
"antlr4.error.ErrorListener",
warn_not_installed=True,
import_kwargs={"fromlist": ["ErrorListener"]},
"""Tests that the IPython printing module is properly loaded. """
from sympy.interactive.session import init_ipython_session
from sympy.external import import_module
ipython = import_module("IPython", min_module_version="0.11")
# disable tests if ipython is not present
if not ipython:
disabled = True
def test_ipythonprinting():
# Initialize and setup IPython session
app = init_ipython_session()
app.run_cell("ip = get_ipython()")
app.run_cell("inst = ip.instance()")
app.run_cell("format = inst.display_formatter.format")
app.run_cell("from sympy import Symbol")
# Printing without printing extension
app.run_cell("a = format(Symbol('pi'))")
assert app.user_ns['a']['text/plain'] == "pi"
# Load printing extension
app.run_cell("%load_ext sympy.interactive.ipythonprinting")
# Printing with printing extension
app.run_cell("a = format(Symbol('pi'))")
assert app.user_ns['a']['text/plain'] == u'\u03c0'
from sympy.utilities.pytest import XFAIL, raises
from sympy import (symbols, lambdify, sqrt, sin, cos, pi, Rational, Float,
Matrix, Lambda, exp, Integral, oo, I, Abs)
from sympy.printing.lambdarepr import LambdaPrinter
from sympy import mpmath
from sympy.utilities.lambdify import implemented_function
from sympy.utilities.pytest import skip
from sympy.utilities.decorator import conserve_mpmath_dps
from sympy.external import import_module
import math
import sympy
MutableDenseMatrix = Matrix
numpy = import_module('numpy', min_python_version=(2, 6))
x, y, z = symbols('x,y,z')
#================== Test different arguments ==============
def test_no_args():
f = lambdify([], 1)
raises(TypeError, lambda: f(-1))
assert f() == 1
def test_single_arg():
f = lambdify(x, 2 * x)
assert f(1) == 2
from sympy import symbols, Eq
from sympy.external import import_module
from sympy.tensor import IndexedBase, Idx
from sympy.utilities.autowrap import autowrap, ufuncify, CodeWrapError
from sympy.utilities.pytest import XFAIL, skip
numpy = import_module('numpy')
Cython = import_module('Cython')
f2py = import_module('numpy.f2py', __import__kwargs={'fromlist': ['f2py']})
f2pyworks = False
if f2py:
try:
autowrap(symbols('x'), 'f95', 'f2py')
except (CodeWrapError, ImportError, OSError):
f2pyworks = False
else:
f2pyworks = True
a, b, c = symbols('a b c')
n, m, d = symbols('n m d', integer=True)
A, B, C = symbols('A B C', cls=IndexedBase)
i = Idx('i', m)
j = Idx('j', n)
k = Idx('k', d)
def has_module(module):
"""
Return True if module exists, otherwise run skip().
from __future__ import unicode_literals, print_function
from sympy.external import import_module
import os
cin = import_module('clang.cindex', import_kwargs = {'fromlist': ['cindex']})
"""
This module contains all the necessary Classes and Function used to Parse C and
C++ code into SymPy expression
The module serves as a backend for SymPyExpression to parse C code
It is also dependent on Clang's AST and Sympy's Codegen AST.
The module only supports the features currently supported by the Clang and
codegen AST which will be updated as the development of codegen AST and this
module progresses.
You might find unexpected bugs and exceptions while using the module, feel free
to report them to the SymPy Issue Tracker
Features Supported
==================
- Variable Declarations (integers and reals)
- Assignment (using integer & floating literal and function calls)
- Function Definitions nad Declaration
- Function Calls
- Compound statements, Return statements
Notes
=====
The module is dependent on an external dependency which needs to be installed
to use the features of this module.
from sympy.external import import_module
if import_module('llvmlite'):
import sympy.printing.llvmjitcode as g
else:
disabled = True
import sympy
from sympy.abc import a, b
# copied from numpy.isclose documentation
def isclose(a, b):
rtol = 1e-5
atol = 1e-8
return abs(a - b) <= atol + rtol * abs(b)
def test_simple_expr():
e = a + 1.0
f = g.llvm_callable([a], e)
res = float(e.subs({a: 4.0}).evalf())
jit_res = f(4.0)
assert isclose(jit_res, res)
def test_two_arg():
e = 4.0 * a + b + 3.0
f = g.llvm_callable([a, b], e)
res = float(e.subs({a: 4.0, b: 3.0}).evalf())
import sys
from sympy.external import import_module
autolevparser = import_module('sympy.parsing.autolev._antlr.autolevparser',
__import__kwargs={'fromlist': ['AutolevParser']})
autolevlexer = import_module('sympy.parsing.autolev._antlr.autolevlexer',
__import__kwargs={'fromlist': ['AutolevLexer']})
autolevlistener = import_module(
'sympy.parsing.autolev._antlr.autolevlistener',
__import__kwargs={'fromlist': ['AutolevListener']})
AutolevParser = getattr(autolevparser, 'AutolevParser', None)
AutolevLexer = getattr(autolevlexer, 'AutolevLexer', None)
AutolevListener = getattr(autolevlistener, 'AutolevListener', None)
def parse_autolev(autolev_code, include_numeric):
antlr4 = import_module('antlr4', warn_not_installed=True)
if not antlr4:
raise ImportError("Autolev parsing requires the antlr4 python package,"
" provided by pip (antlr4-python2-runtime or"
" antlr4-python3-runtime) or"
" conda (antlr-python-runtime)")
try:
l = autolev_code.readlines()
input_stream = antlr4.InputStream("".join(l))
except Exception:
input_stream = antlr4.InputStream(autolev_code)
if AutolevListener:
from ._listener_autolev_antlr import MyListener
from sympy.testing.pytest import raises
from sympy.parsing.sym_expr import SymPyExpression
from sympy.external import import_module
lfortran = import_module('lfortran')
if lfortran:
from sympy.codegen.ast import (Variable, IntBaseType, FloatBaseType,
String, Return, FunctionDefinition,
Assignment, Declaration, CodeBlock)
from sympy.core import Integer, Float, Add
from sympy import Symbol
expr1 = SymPyExpression()
expr2 = SymPyExpression()
src = """\
integer :: a, b, c, d
real :: p, q, r, s
"""
def test_sym_expr():
src1 = (src + """\
d = a + b -c
""")
expr3 = SymPyExpression(src, 'f')
expr4 = SymPyExpression(src1, 'f')
ls1 = expr3.return_expr()
ls2 = expr4.return_expr()
for i in range(0, 7):
assert isinstance(ls1[i], Declaration)
assert isinstance(ls2[i], Declaration)
def _get_raster_interval(self, func):
""" Uses interval math to adaptively mesh and obtain the plot"""
k = self.depth
interval_list = []
#Create initial 32 divisions
np = import_module('numpy')
xsample = np.linspace(self.start_x, self.end_x, 33)
ysample = np.linspace(self.start_y, self.end_y, 33)
#Add a small jitter so that there are no false positives for equality.
# Ex: y==x becomes True for x interval(1, 2) and y interval(1, 2)
#which will draw a rectangle.
jitterx = (np.random.rand(
len(xsample)) * 2 - 1) * (self.end_x - self.start_x) / 2**20
jittery = (np.random.rand(
len(ysample)) * 2 - 1) * (self.end_y - self.start_y) / 2**20
xsample += jitterx
ysample += jittery
xinter = [interval(x1, x2) for x1, x2 in zip(xsample[:-1],
xsample[1:])]
yinter = [interval(y1, y2) for y1, y2 in zip(ysample[:-1],
ysample[1:])]
interval_list = [[x, y] for x in xinter for y in yinter]
plot_list = []
#recursive call refinepixels which subdivides the intervals which are
#neither True nor False according to the expression.
def refine_pixels(interval_list):
""" Evaluates the intervals and subdivides the interval if the
expression is partially satisfied."""
temp_interval_list = []
plot_list = []
for intervals in interval_list:
#Convert the array indices to x and y values
intervalx = intervals[0]
intervaly = intervals[1]
func_eval = func(intervalx, intervaly)
#The expression is valid in the interval. Change the contour
#array values to 1.
if func_eval[1] is False or func_eval[0] is False:
pass
elif func_eval == (True, True):
plot_list.append([intervalx, intervaly])
elif func_eval[1] is None or func_eval[0] is None:
#Subdivide
avgx = intervalx.mid
avgy = intervaly.mid
a = interval(intervalx.start, avgx)
b = interval(avgx, intervalx.end)
c = interval(intervaly.start, avgy)
d = interval(avgy, intervaly.end)
temp_interval_list.append([a, c])
temp_interval_list.append([a, d])
temp_interval_list.append([b, c])
temp_interval_list.append([b, d])
return temp_interval_list, plot_list
while k >= 0 and len(interval_list):
interval_list, plot_list_temp = refine_pixels(interval_list)
plot_list.extend(plot_list_temp)
k = k - 1
#Check whether the expression represents an equality
#If it represents an equality, then none of the intervals
#would have satisfied the expression due to floating point
#differences. Add all the undecided values to the plot.
if self.has_equality:
for intervals in interval_list:
intervalx = intervals[0]
intervaly = intervals[1]
func_eval = func(intervalx, intervaly)
if func_eval[1] and func_eval[0] is not False:
plot_list.append([intervalx, intervaly])
return plot_list, 'fill'
from __future__ import print_function, division
import inspect
from sympy.external import import_module
from sympy.printing.printer import Printer
from sympy.core.compatibility import range
import sympy
from functools import partial
theano = import_module('theano')
if theano:
ts = theano.scalar
tt = theano.tensor
from theano.sandbox import linalg as tlinalg
mapping = {
sympy.Add: tt.add,
sympy.Mul: tt.mul,
sympy.Abs: tt.abs_,
sympy.sign: tt.sgn,
sympy.ceiling: tt.ceil,
sympy.floor: tt.floor,
sympy.log: tt.log,
sympy.exp: tt.exp,
sympy.sqrt: tt.sqrt,
sympy.cos: tt.cos,
sympy.acos: tt.arccos,
sympy.sin: tt.sin,
sympy.asin: tt.arcsin,
sympy.tan: tt.tan,
def test_empty_plot():
matplotlib = import_module('matplotlib', min_module_version='1.1.0', catch=(RuntimeError,))
if not matplotlib:
skip("Matplotlib not the default backend")
# No exception showing an empty plot
plot()
from sympy.printing.lambdarepr import LambdaPrinter
from sympy import mpmath
from sympy.utilities.lambdify import implemented_function
from sympy.utilities.pytest import skip
from sympy.utilities.decorator import conserve_mpmath_dps
from sympy.external import import_module
import math
import sympy
# TODO: This should be removed for the release of 0.7.7, see issue #7853
from functools import partial
lambdify = partial(lambdify, default_array=True)
MutableDenseMatrix = Matrix
numpy = import_module('numpy')
numexpr = import_module('numexpr')
w, x, y, z = symbols('w,x,y,z')
#================== Test different arguments =======================
def test_no_args():
f = lambdify([], 1)
raises(TypeError, lambda: f(-1))
assert f() == 1
def test_single_arg():
f = lambdify(x, 2*x)
def test_sampling_gamma_inverse():
scipy = import_module('scipy')
if not scipy:
skip('Scipy not installed. Abort tests for sampling of gamma inverse.')
X = GammaInverse("x", 1, 1)
assert sample(X) in X.pspace.domain.set
import tempfile
import sympy as sp
from sympy.codegen.ast import Assignment
from sympy.codegen.algorithms import newtons_method, newtons_method_function
from sympy.codegen.fnodes import bind_C
from sympy.codegen.futils import render_as_module as f_module
from sympy.codegen.pyutils import render_as_module as py_module
from sympy.external import import_module
from sympy.printing import ccode
from sympy.utilities._compilation import compile_link_import_strings, has_c, has_fortran
from sympy.utilities._compilation.util import may_xfail
from sympy.testing.pytest import skip, raises
cython = import_module('cython')
wurlitzer = import_module('wurlitzer')
def test_newtons_method():
x, dx, atol = sp.symbols('x dx atol')
expr = sp.cos(x) - x**3
algo = newtons_method(expr, x, atol, dx)
assert algo.has(Assignment(dx, -expr / expr.diff(x)))
@may_xfail
def test_newtons_method_function__ccode():
x = sp.Symbol('x', real=True)
expr = sp.cos(x) - x**3
func = newtons_method_function(expr, x)
if not cython:
def __call__(self, *args):
np = import_module("numpy")
np_old_err = np.seterr(invalid="raise")
try:
temp_args = (np.array(a, dtype=np.complex) for a in args)
results = self.vector_func(*temp_args)
results = np.ma.masked_where(
np.abs(results.imag) > 1e-7 * np.abs(results),
results.real,
copy=False)
except Exception as e:
# DEBUG: print 'Error', type(e), e
if (isinstance(e, TypeError)
and "unhashable type: 'numpy.ndarray'" in str(e)) or (
isinstance(e, ValueError) and
("Invalid limits given:" in str(e) or
"negative dimensions are not allowed" in str(e) # XXX
or "sequence too large; must be smaller than 32"
in str(e))): # XXX
# Almost all functions were translated to numpy, but some were
# left as sympy functions. They received an ndarray as an
# argument and failed.
# sin(ndarray(...)) raises "unhashable type"
# Integral(x, (x, 0, ndarray(...))) raises "Invalid limits"
# other ugly exceptions that are not well understood (marked with XXX)
# TODO: Cleanup the ugly special cases marked with xxx above.
# Solution: use cmath and vectorize the final lambda.
self.lambda_func = experimental_lambdify(self.args,
self.expr,
use_python_cmath=True)
self.vector_func = np.vectorize(self.lambda_func,
otypes=[np.complex])
results = self.vector_func(*args)
results = np.ma.masked_where(
np.abs(results.imag) > 1e-7 * np.abs(results),
results.real,
copy=False,
)
else:
# Complete failure. One last try with no translations, only
# wrapping in complex((...).evalf()) and returning the real
# part.
if self.failure:
raise e
else:
self.failure = True
self.lambda_func = experimental_lambdify(
self.args,
self.expr,
use_evalf=True,
complex_wrap_evalf=True)
self.vector_func = np.vectorize(self.lambda_func,
otypes=[np.complex])
results = self.vector_func(*args)
results = np.ma.masked_where(
np.abs(results.imag) > 1e-7 * np.abs(results),
results.real,
copy=False,
)
warnings.warn(
"The evaluation of the expression is"
" problematic. We are trying a failback method"
" that may still work. Please report this as a bug.")
finally:
np.seterr(**np_old_err)
return results