Evaluating a String in Sage#

sage.misc.sage_eval.sage_eval(source, locals=None, cmds='', preparse=True)#

Obtain a Sage object from the input string by evaluating it using Sage. This means calling eval after preparsing and with globals equal to everything included in the scope of from sage.all import *.).


  • source - a string or object with a _sage_ method

  • locals - evaluate in namespace of sage.all plus the locals dictionary

  • cmds - string; sequence of commands to be run before source is evaluated.

  • preparse - (default: True) if True, preparse the string expression.

EXAMPLES: This example illustrates that preparsing is applied.

sage: eval('2^3')
sage: sage_eval('2^3')

However, preparsing can be turned off.

sage: sage_eval('2^3', preparse=False)

Note that you can explicitly define variables and pass them as the second option:

sage: x = PolynomialRing(RationalField(),"x").gen()
sage: sage_eval('x^2+1', locals={'x':x})
x^2 + 1

This example illustrates that evaluation occurs in the context of from sage.all import *. Even though bernoulli has been redefined in the local scope, when calling sage_eval() the default value meaning of bernoulli() is used. Likewise for QQ below.

sage: bernoulli = lambda x : x^2
sage: bernoulli(6)
sage: eval('bernoulli(6)')
sage: sage_eval('bernoulli(6)')                                                 # needs sage.libs.flint
sage: QQ = lambda x : x^2
sage: QQ(2)
sage: sage_eval('QQ(2)')
sage: parent(sage_eval('QQ(2)'))
Rational Field

This example illustrates setting a variable for use in evaluation.

sage: x = 5
sage: eval('4//3 + x', {'x': 25})
sage: sage_eval('4/3 + x',  locals={'x': 25})

You can also specify a sequence of commands to be run before the expression is evaluated:

sage: sage_eval('p', cmds='K.<x> = QQ[]\np = x^2 + 1')
x^2 + 1

If you give commands to execute and a dictionary of variables, then the dictionary will be modified by assignments in the commands:

sage: vars = {}
sage: sage_eval('None', cmds='y = 3', locals=vars)
sage: vars['y'], parent(vars['y'])
(3, Integer Ring)

You can also specify the object to evaluate as a tuple. A 2-tuple is assumed to be a pair of a command sequence and an expression; a 3-tuple is assumed to be a triple of a command sequence, an expression, and a dictionary holding local variables. (In this case, the given dictionary will not be modified by assignments in the commands.)

sage: sage_eval(('f(x) = x^2', 'f(3)'))                                         # needs sage.symbolic
sage: vars = {'rt2': sqrt(2.0)}
sage: sage_eval(('rt2 += 1', 'rt2', vars))
sage: vars['rt2']

This example illustrates how sage_eval can be useful when evaluating the output of other computer algebra systems.

sage: # needs sage.libs.gap
sage: R.<x> = PolynomialRing(RationalField())
sage: gap.eval('R:=PolynomialRing(Rationals,["x"]);')
sage: ff = gap.eval('x:=IndeterminatesOfPolynomialRing(R);; f:=x^2+1;'); ff
sage: sage_eval(ff, locals={'x':x})
x^2 + 1
sage: eval(ff)
Traceback (most recent call last):
RuntimeError: Use ** for exponentiation, not '^', which means xor
in Python, and has the wrong precedence.

Here you can see that eval() simply will not work but sage_eval() will.

sage.misc.sage_eval.sageobj(x, vars=None)#

Return a native Sage object associated to x, if possible and implemented.

If the object has a _sage_ method it is called and the value is returned. Otherwise, str() is called on the object, and all preparsing is applied and the resulting expression is evaluated in the context of from sage.all import *. To evaluate the expression with certain variables set, use the vars argument, which should be a dictionary.


sage: type(sageobj(gp('34/56')))                                                # needs sage.libs.pari
<class 'sage.rings.rational.Rational'>

sage: n = 5/2
sage: sageobj(n) is n
sage: k = sageobj('Z(8^3/1)', {'Z':ZZ}); k
sage: type(k)
<class 'sage.rings.integer.Integer'>

This illustrates interfaces:

sage: # needs sage.libs.pari
sage: f = gp('2/3')
sage: type(f)
<class 'sage.interfaces.gp.GpElement'>
sage: f._sage_()
sage: type(f._sage_())
<class 'sage.rings.rational.Rational'>

sage: # needs sage.libs.gap
sage: a = gap(939393/2433)
sage: a._sage_()
sage: type(a._sage_())
<class 'sage.rings.rational.Rational'>