The problem is that is_lowercase looks like the assumptions system. So
people might become confused when the assumptions APIs don't actually
work. For instance
>>> Symbol('x', number=True).is_number
False
At least is_polynomial and is_rational_function are functions.
I think the fix here is to make the above emit a warning (see
https://github.com/sympy/sympy/pull/21417#issuecomment-856001694). And
everything should be documented as well.
Aaron Meurer
On Mon, Sep 13, 2021 at 6:40 PM Chris Smith <[email protected]> wrote:
>
> The varieties of `is_lowercase` don't bother me as much: some are easy to
> compute independently (like `is_number`) whereas others are more complicated
> and depend on other attributes (like `is_real`).
>
> /c
>
> On Monday, September 13, 2021 at 1:22:38 PM UTC-5 [email protected] wrote:
>>
>> A glossary would be a great idea.
>>
>> By the way, I am going to start writing a lot of documentation in
>> SymPy soon (see https://groups.google.com/g/sympy/c/vYsavewGj1w). If
>> anyone has things in SymPy that they wish were documented better,
>> particularly the sorts of things that would go in high-level
>> explanations and how-to guides, please reach out to me. We will likely
>> be running some sort of survey for the same so watch this space for
>> more information in the coming months.
>>
>> On Mon, Sep 13, 2021 at 12:14 PM Oscar Benjamin
>> <[email protected]> wrote:
>> >
>> > There is also is_Number vs is_number. There is also is_comparable which
>> > means can be evalf'd to a real number.
>>
>> As an aside, I really wish we didn't have the is_Uppercase attributes.
>> They were implemented as a performance hack because attribute lookups
>> are slightly faster than isinstance(). But the cost they have had in
>> terms of user confusion greatly outweighs any performance gain they
>> have produced. We should consider deprecating them, or at the very
>> least, banning the creation of new ones on new classes. I think it's
>> also unfortunate that we have is_ methods that aren't actually part of
>> the assumptions system. But that's just something we'll just have to
>> document and live with.
>>
>> Aaron Meurer
>>
>> >
>> > There should be a glossary of all the different attributes that are used
>> > throughout sympy:
>> >
>> > In [1]: for name in dir(x):
>> >
>> > ...: if name.startswith('is'):
>> >
>> > ...: print(name)
>> >
>> > ...:
>> >
>> > is_Add
>> >
>> > is_AlgebraicNumber
>> >
>> > is_Atom
>> >
>> > is_Boolean
>> >
>> > is_Derivative
>> >
>> > is_Dummy
>> >
>> > is_Equality
>> >
>> > is_Float
>> >
>> > is_Function
>> >
>> > is_Indexed
>> >
>> > is_Integer
>> >
>> > is_MatAdd
>> >
>> > is_MatMul
>> >
>> > is_Matrix
>> >
>> > is_Mul
>> >
>> > is_Not
>> >
>> > is_Number
>> >
>> > is_NumberSymbol
>> >
>> > is_Order
>> >
>> > is_Piecewise
>> >
>> > is_Point
>> >
>> > is_Poly
>> >
>> > is_Pow
>> >
>> > is_Rational
>> >
>> > is_Relational
>> >
>> > is_Symbol
>> >
>> > is_Vector
>> >
>> > is_Wild
>> >
>> > is_algebraic
>> >
>> > is_algebraic_expr
>> >
>> > is_antihermitian
>> >
>> > is_commutative
>> >
>> > is_comparable
>> >
>> > is_complex
>> >
>> > is_composite
>> >
>> > is_constant
>> >
>> > is_even
>> >
>> > is_extended_negative
>> >
>> > is_extended_nonnegative
>> >
>> > is_extended_nonpositive
>> >
>> > is_extended_nonzero
>> >
>> > is_extended_positive
>> >
>> > is_extended_real
>> >
>> > is_finite
>> >
>> > is_hermitian
>> >
>> > is_hypergeometric
>> >
>> > is_imaginary
>> >
>> > is_infinite
>> >
>> > is_integer
>> >
>> > is_irrational
>> >
>> > is_meromorphic
>> >
>> > is_negative
>> >
>> > is_noninteger
>> >
>> > is_nonnegative
>> >
>> > is_nonpositive
>> >
>> > is_nonzero
>> >
>> > is_number
>> >
>> > is_odd
>> >
>> > is_polar
>> >
>> > is_polynomial
>> >
>> > is_positive
>> >
>> > is_prime
>> >
>> > is_rational
>> >
>> > is_rational_function
>> >
>> > is_real
>> >
>> > is_scalar
>> >
>> > is_symbol
>> >
>> > is_transcendental
>> >
>> > is_zero
>> >
>> >
>> >
>> > On Mon, 13 Sept 2021 at 18:56, Aaron Meurer <[email protected]> wrote:
>> >>
>> >> This would be a good thing to have some standalone documentation on.
>> >> There are also some related things like the functionality that is used
>> >> by diff() to determine what can be used as a differentiation variable,
>> >> which is a little more complex than just free_symbols in general
>> >> because you can have things like derivatives with respect to indexed
>> >> expressions.
>> >>
>> >> Aaron Meurer
>> >>
>> >> On Mon, Sep 13, 2021 at 11:48 AM Chris Smith <[email protected]> wrote:
>> >> >
>> >> > > can be evalf'ed
>> >> >
>> >> > That's a clear and good reminder.
>> >> >
>> >> > /c
>> >> >
>> >> > On Monday, September 13, 2021 at 12:10:50 PM UTC-5 [email protected]
>> >> > wrote:
>> >> >>
>> >> >> is_number means "can be evalf'ed". So for example, we have the
>> >> >> following
>> >> >>
>> >> >> >>> f = Function('f')
>> >> >> >>> f(0).is_number
>> >> >> False
>> >> >> >>> f(0).free_symbols
>> >> >> set()
>> >> >>
>> >> >> So you should use is_number specifically if you are checking if you
>> >> >> can evaluate the expression to a literal number.
>> >> >>
>> >> >> Aaron Meurer
>> >> >>
>> >> >> On Mon, Sep 13, 2021 at 10:21 AM Paul Royik <[email protected]>
>> >> >> wrote:
>> >> >> >
>> >> >> > Thanks to everybody!
>> >> >> >
>> >> >> > On Monday, September 13, 2021 at 3:56:47 PM UTC+3 Oscar wrote:
>> >> >> >>
>> >> >> >> Think about things that are literally not numbers:
>> >> >> >>
>> >> >> >> In [9]: Interval(1, 2).is_number
>> >> >> >> Out[9]: False
>> >> >> >>
>> >> >> >> In [10]: ImmutableMatrix([[1, 2], [3, 4]]).is_number
>> >> >> >> Out[10]: False
>> >> >> >>
>> >> >> >>
>> >> >> >> On Mon, 13 Sept 2021 at 13:00, Chris Smith <[email protected]> wrote:
>> >> >> >>>
>> >> >> >>> To confirm, if you mean that it is free from any Symbol (free or
>> >> >> >>> bound) then `not expr.has(Symbol)` will be best. But if you
>> >> >> >>> consider `Integral(x, (x, 1, 2))` as a number then you should use
>> >> >> >>> `is_number` or `free_symbols`, with `expr.is_number` failing
>> >> >> >>> sooner than `not expr.free_symbols` if the expression has a free
>> >> >> >>> symbol. (So if you suspect the expression has free symbols then
>> >> >> >>> use `is_number`, else `free_symbols`).
>> >> >> >>>
>> >> >> >>> `f.is_number != (not bool(f.free_symbols))` should be an invariant
>> >> >> >>> for Expr, but SymPy also deals with Booleans, so
>> >> >> >>> `S.true.is_number` is False and `S.true.free_symbols` is empty.
>> >> >> >>>
>> >> >> >>> /c
>> >> >> >>>
>> >> >> >>> On Sunday, September 12, 2021 at 11:56:23 PM UTC-5
>> >> >> >>> [email protected] wrote:
>> >> >> >>>>
>> >> >> >>>> Are there any cases when f.is_number != (not
>> >> >> >>>> bool(f.free_symbols))?
>> >> >> >>>>
>> >> >> >>>> If I have an arbitrary expression, what is the correct way to
>> >> >> >>>> check whether it has variables?
>> >> >> >>>>
>> >> >> >>>> Thank you.
>> >> >> >>>
>> >> >> >>> --
>> >> >> >>> You received this message because you are subscribed to the Google
>> >> >> >>> Groups "sympy" group.
>> >> >> >>> To unsubscribe from this group and stop receiving emails from it,
>> >> >> >>> send an email to [email protected].
>> >> >> >>> To view this discussion on the web visit
>> >> >> >>> https://groups.google.com/d/msgid/sympy/9af46205-ce22-494c-a604-c27b6682fa96n%40googlegroups.com.
>> >> >> >
>> >> >> > --
>> >> >> > You received this message because you are subscribed to the Google
>> >> >> > Groups "sympy" group.
>> >> >> > To unsubscribe from this group and stop receiving emails from it,
>> >> >> > send an email to [email protected].
>> >> >> > To view this discussion on the web visit
>> >> >> > https://groups.google.com/d/msgid/sympy/8717be3c-0286-4741-900c-94e21573f3c5n%40googlegroups.com.
>> >> >
>> >> > --
>> >> > You received this message because you are subscribed to the Google
>> >> > Groups "sympy" group.
>> >> > To unsubscribe from this group and stop receiving emails from it, send
>> >> > an email to [email protected].
>> >> > To view this discussion on the web visit
>> >> > https://groups.google.com/d/msgid/sympy/69d8097b-b81b-4490-ae85-0dbb946346c1n%40googlegroups.com.
>> >>
>> >> --
>> >> You received this message because you are subscribed to the Google Groups
>> >> "sympy" group.
>> >> To unsubscribe from this group and stop receiving emails from it, send an
>> >> email to [email protected].
>> >> To view this discussion on the web visit
>> >> https://groups.google.com/d/msgid/sympy/CAKgW%3D6LOouUDFL9eooBukJz5G%2BSSMOoFWkVBLvwbqjrbduSjkA%40mail.gmail.com.
>> >
>> > --
>> > You received this message because you are subscribed to the Google Groups
>> > "sympy" group.
>> > To unsubscribe from this group and stop receiving emails from it, send an
>> > email to [email protected].
>> > To view this discussion on the web visit
>> > https://groups.google.com/d/msgid/sympy/CAHVvXxQeJ8eOdCpSKsiAXPjZKrd_jS2WRmfLdZS_Drjm377L5g%40mail.gmail.com.
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sympy/50f5774b-7835-470c-8a08-57617910ec4en%40googlegroups.com.
--
You received this message because you are subscribed to the Google Groups
"sympy" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/sympy/CAKgW%3D6LggK6hDEZFyn%2BSmGNHh99VD%2BG_eRd8mPHWS0u7foR1EA%40mail.gmail.com.
