maxima-bugs — Reports from the SF bug-tracking system

[Maxima-bugs] [maxima:bugs] #4590 integrals of some rational functions from minf to inf

[Barton W. <wil...@us...>]

---

**[bugs:#4590] integrals of some rational functions from minf to inf**

**Status:** open
**Group:** None
**Labels:** zmtorat 
**Created:** Sun Aug 03, 2025 06:13 PM UTC by Barton Willis
**Last Updated:** Sun Aug 03, 2025 06:13 PM UTC
**Owner:** nobody


This is OK
~~~
(%i15) integrate(1/(x^2 + %i+1),x,minf,inf);
(%o15) -((sqrt(sqrt(2)+1)*((sqrt(2)-1)*%i-1)*%pi)/2)
~~~
But this is  incorrect and inconsistent with `%o15`
~~~
(%i16) integrate(1/(x^2 + %i+a),x,minf,inf);
(%o16) 0
~~~
Both of these functions call  `zmtorat`


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4589 guess fails on trivial cases

[Stavros M. <mac...@us...>]

Ah! The documentation is actually at a different place (not devine): /share/contrib/guess/guess.info

That clarifies that "product" here means not multiplication, but an indexed product(...). Which also explains how it handles 2^n = product(2,i,1,n) and n! = product(i,i,1,n).


---

**[bugs:#4589] guess fails on trivial cases**

**Status:** open
**Group:** None
**Created:** Wed Jul 30, 2025 04:39 PM UTC by Stavros Macrakis
**Last Updated:** Wed Jul 30, 2025 08:49 PM UTC
**Owner:** nobody


~~~~
load(devine)$
guess([0,0,0]) => expt: undefined: 0 to a negative exponent.
guess([0,1,0,1]) => same error
guess(makelist(1+2^i,i,1,10)) => fails
~~~~

The only documentation for this function is in the source, and it says cryptically 
> it tries to find a closed form for a sequence within the hierarchy of expressions of the form <rational function\>, <product of rational functions>, <product of product of rational functions>, etc. It may give several answers 

Ah, so those sequences (not even 0,0,0?) aren't in its domain.

Also, it appears that the required length of the sequence grows with the exponents -- this should be documented:
~~~~
guess(makelist(i^9,i,1,10)) => []
guess(makelist(i^9,i,1,11)) => [i0^9] <<< OK!
guess(makelist(i^9,i,1,20)) => Extremely slow
guess(makelist(i^5/(i^5+1),i,1,10)) => []    
guess(makelist(i^5/(i^5+1),i,1,14)) => OK
~~~~
At the same time, it looks like time is super-exponential in the length of the sequence.

And surely ''guess'' is an awfully broad name for such a special-purpose function!

5.48.0 SBCL 2.5.7 MacOS


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4589 guess fails on trivial cases

[Stavros M. <mac...@us...>]

The doc is also bizarre because on the one hand, a product of rational functions is itself a rational function, and on the other hand, it handles cases which aren't listed there, like geometric series:
~~~~
guess([1,1/2,1/4,1/8])
     => [2^(1-i0),2^(1-i0)]
 ~~~~
 Each item in the return list is supposed to be a different expression, but here they're all the same.
 ~~~~
 guess(makelist(2^i,i,0,10)) =>
    [2^(i0-1),2^(i0-1),2^(i0-1),2^(i0-1),2^(i0-1),2^(i0-1),2^(i0-1),2^(i0-1),2^(i0-1)]
~~~~

 


---

**[bugs:#4589] guess fails on trivial cases**

**Status:** open
**Group:** None
**Created:** Wed Jul 30, 2025 04:39 PM UTC by Stavros Macrakis
**Last Updated:** Wed Jul 30, 2025 04:39 PM UTC
**Owner:** nobody


~~~~
load(devine)$
guess([0,0,0]) => expt: undefined: 0 to a negative exponent.
guess([0,1,0,1]) => same error
guess(makelist(1+2^i,i,1,10)) => fails
~~~~

The only documentation for this function is in the source, and it says cryptically 
> it tries to find a closed form for a sequence within the hierarchy of expressions of the form <rational function\>, <product of rational functions>, <product of product of rational functions>, etc. It may give several answers 

Ah, so those sequences (not even 0,0,0?) aren't in its domain.

Also, it appears that the required length of the sequence grows with the exponents -- this should be documented:
~~~~
guess(makelist(i^9,i,1,10)) => []
guess(makelist(i^9,i,1,11)) => [i0^9] <<< OK!
guess(makelist(i^9,i,1,20)) => Extremely slow
guess(makelist(i^5/(i^5+1),i,1,10)) => []    
guess(makelist(i^5/(i^5+1),i,1,14)) => OK
~~~~
At the same time, it looks like time is super-exponential in the length of the sequence.

And surely ''guess'' is an awfully broad name for such a special-purpose function!

5.48.0 SBCL 2.5.7 MacOS


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4589 guess fails on trivial cases

[Stavros M. <mac...@us...>]

---

**[bugs:#4589] guess fails on trivial cases**

**Status:** open
**Group:** None
**Created:** Wed Jul 30, 2025 04:39 PM UTC by Stavros Macrakis
**Last Updated:** Wed Jul 30, 2025 04:39 PM UTC
**Owner:** nobody


~~~~
load(devine)$
guess([0,0,0]) => expt: undefined: 0 to a negative exponent.
guess([0,1,0,1]) => same error
guess(makelist(1+2^i,i,1,10)) => fails
~~~~

The only documentation for this function is in the source, and it says cryptically 
> it tries to find a closed form for a sequence within the hierarchy of expressions of the form <rational function\>, <product of rational functions>, <product of product of rational functions>, etc. It may give several answers 

Ah, so those sequences (not even 0,0,0?) aren't in its domain.

Also, it appears that the required length of the sequence grows with the exponents -- this should be documented:
~~~~
guess(makelist(i^9,i,1,10)) => []
guess(makelist(i^9,i,1,11)) => [i0^9] <<< OK!
guess(makelist(i^9,i,1,20)) => Extremely slow
guess(makelist(i^5/(i^5+1),i,1,10)) => []    
guess(makelist(i^5/(i^5+1),i,1,14)) => OK
~~~~
At the same time, it looks like time is super-exponential in the length of the sequence.

And surely ''guess'' is an awfully broad name for such a special-purpose function!

5.48.0 SBCL 2.5.7 MacOS


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4587 limit of integral expression incorrect, should be 1/e^2, not 0

[Raymond T. <rt...@us...>]

- Description has changed:

Diff:

~~~~

--- old
+++ new
@@ -1,3 +1,4 @@
+```
 (%i40)	n(x) := integrate((1-1/t)^t*%e^(%e*t), t, %e, x);
 	n(%e+1);
 	d(x) := %e^(%e*x);
@@ -6,3 +7,4 @@
 (%o38)	integrate((1-1/t)^t*%e^(%e*t),t,%e,%e+1)
 (%o39)	d(x):=%e^(%e*x)
 (%o40)	0
+```

~~~~




---

**[bugs:#4587] limit of integral expression incorrect, should be 1/e^2, not 0**

**Status:** open
**Group:** None
**Created:** Wed Jul 30, 2025 07:56 AM UTC by vsxbamboo
**Last Updated:** Wed Jul 30, 2025 03:10 PM UTC
**Owner:** nobody


```
(%i40)	n(x) := integrate((1-1/t)^t*%e^(%e*t), t, %e, x);
	n(%e+1);
	d(x) := %e^(%e*x);
	limit(n(x)/d(x), x, inf);
(%o37)	n(x):=integrate((1-1/t)^t*%e^(%e*t),t,%e,x)
(%o38)	integrate((1-1/t)^t*%e^(%e*t),t,%e,%e+1)
(%o39)	d(x):=%e^(%e*x)
(%o40)	0
```


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4587 limit of integral expression incorrect, should be 1/e^2, not 0

[Stavros M. <mac...@us...>]

I think the OP did in fact include multiplication signs. But sourceforge interpreted them as formatting. Compare this (in text mode):
: a*b^3*5
with this (in code mode):
~~~~
a*b^3*5
~~~~
The text I entered is the same in both cases, but in the first, the asterisk is interpreted as "start italics". As far as I can tell, there isn't any way to see the OP's raw input, but you can look at the italic transitions and reconstruct their input as 
~~~~
n(x) := integrate((1-1/t)^t*%e^(%e*t), t, %e, x);
~~~~
which is what you suggested.

I don't know how we can help users get this right....


---

**[bugs:#4587] limit of integral expression incorrect, should be 1/e^2, not 0**

**Status:** open
**Group:** None
**Created:** Wed Jul 30, 2025 07:56 AM UTC by vsxbamboo
**Last Updated:** Wed Jul 30, 2025 09:41 AM UTC
**Owner:** nobody


(%i40)	n(x) := integrate((1-1/t)^t*%e^(%e*t), t, %e, x);
	n(%e+1);
	d(x) := %e^(%e*x);
	limit(n(x)/d(x), x, inf);
(%o37)	n(x):=integrate((1-1/t)^t*%e^(%e*t),t,%e,x)
(%o38)	integrate((1-1/t)^t*%e^(%e*t),t,%e,%e+1)
(%o39)	d(x):=%e^(%e*x)
(%o40)	0


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4587 limit of integral expression incorrect, should be 1/e^2, not 0

[Barton W. <wil...@us...>]

Maybe your code as shown shows a bug, but I'm wondering if  your input is missing a few explicit multiplication signs? Maybe the definition of `n` should be something like 
~~~
n(x) := integrate((1-1/t)^t * %e^(%e * t), t, %e, x);
~~~




---

**[bugs:#4587] limit of integral expression incorrect, should be 1/e^2, not 0**

**Status:** open
**Group:** None
**Created:** Wed Jul 30, 2025 07:56 AM UTC by vsxbamboo
**Last Updated:** Wed Jul 30, 2025 07:56 AM UTC
**Owner:** nobody


(%i40)	n(x) := integrate((1-1/t)^t*%e^(%e*t), t, %e, x);
	n(%e+1);
	d(x) := %e^(%e*x);
	limit(n(x)/d(x), x, inf);
(%o37)	n(x):=integrate((1-1/t)^t*%e^(%e*t),t,%e,x)
(%o38)	integrate((1-1/t)^t*%e^(%e*t),t,%e,%e+1)
(%o39)	d(x):=%e^(%e*x)
(%o40)	0


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4588 simplification of integrands of definite integrals

[Barton W. <wil...@us...>]

---

**[bugs:#4588] simplification of integrands of definite integrals**

**Status:** open
**Group:** None
**Created:** Wed Jul 30, 2025 09:33 AM UTC by Barton Willis
**Last Updated:** Wed Jul 30, 2025 09:33 AM UTC
**Owner:** nobody


In  input 3, Maxima should simplify the integrand to `z`. Then `%o3` would be `1/2`.
~~~
(%i1) display2d : false$

(%i2) declare(z,complex)$

(%i3) integrate(conjugate(z),z,0,1);
(%o3) 'integrate(conjugate(z),z,0,1)
~~~
This is a minor problem, but the testsuite has some variations on this bug.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4587 limit of integral expression incorrect, should be 1/e^2, not 0

[vsxbamboo <vsx...@us...>]

---

**[bugs:#4587] limit of integral expression incorrect, should be 1/e^2, not 0**

**Status:** open
**Group:** None
**Created:** Wed Jul 30, 2025 07:56 AM UTC by vsxbamboo
**Last Updated:** Wed Jul 30, 2025 07:56 AM UTC
**Owner:** nobody


(%i40)	n(x) := integrate((1-1/t)^t*%e^(%e*t), t, %e, x);
	n(%e+1);
	d(x) := %e^(%e*x);
	limit(n(x)/d(x), x, inf);
(%o37)	n(x):=integrate((1-1/t)^t*%e^(%e*t),t,%e,x)
(%o38)	integrate((1-1/t)^t*%e^(%e*t),t,%e,%e+1)
(%o39)	d(x):=%e^(%e*x)
(%o40)	0


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4586 nointegrate undocumented

[Barton W. <wil...@us...>]

I agree--let's not document it and work toward fixing its bugs.  Plus it's weird to tell users that a flag inhibits the use of one internal method--users shouldn't have to know about such internals. 

Currently, the `antideriv` method is protected buy an long list of operator names--I think the only logic in the list of operator names is that somebody found a bug for an expression that involved an operator in the list.  That, unfortunately, eliminates lots of problems that `antideriv` can handle just fine.


---

**[bugs:#4586] nointegrate undocumented**

**Status:** open
**Group:** None
**Labels:** nointegrate 
**Created:** Mon Jul 28, 2025 06:35 PM UTC by Barton Willis
**Last Updated:** Mon Jul 28, 2025 07:17 PM UTC
**Owner:** nobody


The option variable `nointegrate` is undocumented. A source code comment notes:

*nointegrate is a Macsyma-level flag which inhibits indefinite integration.*

It appears that when `nointegrate` is set to true, definite integration does not use the method that calls  antideriv, and nothing more.

I propose that we focus on fixing the bugs in the antideriv code with the eventual goal of eliminating the need for the nointegrate flag. Till then, if somebody wants to document `nointegrate`,  that's OK.





---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4586 nointegrate undocumented

[Raymond T. <rt...@us...>]

I propose that we don't document it, especially if the goal is eliminating the need for it.

Besides, if the user really wants it, they can do the substitution for the limits themselves.  (Assuming the definite integral doesn't come from some internal maxima function.)


---

**[bugs:#4586] nointegrate undocumented**

**Status:** open
**Group:** None
**Labels:** nointegrate 
**Created:** Mon Jul 28, 2025 06:35 PM UTC by Barton Willis
**Last Updated:** Mon Jul 28, 2025 06:35 PM UTC
**Owner:** nobody


The option variable `nointegrate` is undocumented. A source code comment notes:

*nointegrate is a Macsyma-level flag which inhibits indefinite integration.*

It appears that when `nointegrate` is set to true, definite integration does not use the method that calls  antideriv, and nothing more.

I propose that we focus on fixing the bugs in the antideriv code with the eventual goal of eliminating the need for the nointegrate flag. Till then, if somebody wants to document `nointegrate`,  that's OK.





---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4586 nointegrate undocumented

[Barton W. <wil...@us...>]

---

**[bugs:#4586] nointegrate undocumented**

**Status:** open
**Group:** None
**Labels:** nointegrate 
**Created:** Mon Jul 28, 2025 06:35 PM UTC by Barton Willis
**Last Updated:** Mon Jul 28, 2025 06:35 PM UTC
**Owner:** nobody


The option variable `nointegrate` is undocumented. A source code comment notes:

*nointegrate is a Macsyma-level flag which inhibits indefinite integration.*

It appears that when `nointegrate` is set to true, definite integration does not use the method that calls  antideriv, and nothing more.

I propose that we focus on fixing the bugs in the antideriv code with the eventual goal of eliminating the need for the nointegrate flag. Till then, if somebody wants to document `nointegrate`,  that's OK.





---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4581 integrate ((log (1 - t)) / t, t, 0, x) is wrong

[Barton W. <wil...@us...>]

I think the problem  is in the functions `logx1`, `intcv,` and `intcv2`, possibly due to a bogus value of the special variable.  


---

**[bugs:#4581] integrate ((log (1 - t)) / t, t, 0, x) is wrong**

**Status:** open
**Group:** None
**Labels:** integrate 
**Created:** Sun Jul 20, 2025 05:19 PM UTC by Raymond Toy
**Last Updated:** Thu Jul 24, 2025 11:53 AM UTC
**Owner:** nobody


In current maxima, we have
```maxima
(%i1) display2d:false;

(%o1) false
(%i2) integrate ((log (1 - t)) / t, t, 0, x);
Is x positive, negative or zero?

pos;
Is x-1 positive, negative or zero?

neg;

(%o2) (log(1-x)*(2*log(x)-2*log(x-1))+log(1-x)^2-2*li[2](-(1/(x-1))))/2+%pi^2
                                                                        /6
```
According to the user manual, this used to return `-li[2](x)`.

If you differentiate `%o2` and manipulate it, you get the negative of the integrand.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4581 integrate ((log (1 - t)) / t, t, 0, x) is wrong

[Barton W. <wil...@us...>]

...and one more thing: what is the magic in `ldefint` that makes the following correct?

~~~
(%i6) ldefint((log (1 - t)) / t, t,0,x);

INTEGRALLOOKUPS: Found integral MEXPT.
INTEGRALLOOKUPS: Found integral MEXPT.
INTFORM: found 'INTEGRAL on property list
INTEGRALLOOKUPS: Found integral MEXPT.
INTFORM: found 'INTEGRAL on property list
INTEGRALLOOKUPS: Found integral MEXPT.
INTFORM: found 'INTEGRAL on property list
INTEGRALLOOKUPS: Found integral MEXPT.
INTFORM: found 'INTEGRAL on property list
INTEGRALLOOKUPS: Found integral MEXPT.
                                                         2
                                                      %pi
(%o6)                log(1 - x) log(x) + li (1 - x) - ────
                                           2           6
(%i7) limit(%,x,0,plus);
(%o7)                                  0
(%i8) diff(%o6,x);
                                  log(1 - x)
(%o8)                             ──────────
                                      x

~~~


---

**[bugs:#4581] integrate ((log (1 - t)) / t, t, 0, x) is wrong**

**Status:** open
**Group:** None
**Labels:** integrate 
**Created:** Sun Jul 20, 2025 05:19 PM UTC by Raymond Toy
**Last Updated:** Thu Jul 24, 2025 11:50 AM UTC
**Owner:** nobody


In current maxima, we have
```maxima
(%i1) display2d:false;

(%o1) false
(%i2) integrate ((log (1 - t)) / t, t, 0, x);
Is x positive, negative or zero?

pos;
Is x-1 positive, negative or zero?

neg;

(%o2) (log(1-x)*(2*log(x)-2*log(x-1))+log(1-x)^2-2*li[2](-(1/(x-1))))/2+%pi^2
                                                                        /6
```
According to the user manual, this used to return `-li[2](x)`.

If you differentiate `%o2` and manipulate it, you get the negative of the integrand.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4581 integrate ((log (1 - t)) / t, t, 0, x) is wrong

[Barton W. <wil...@us...>]

Maybe this is the same as calling risch, but we also have
~~~
(%i1) integrate ((log (1 - t)) / t, t);

INTEGRALLOOKUPS: Found integral MEXPT.
INTEGRALLOOKUPS: Found integral MEXPT.
INTEGRALLOOKUPS: Found integral MEXPT.
INTEGRALLOOKUPS: Found integral MEXPT.
INTEGRALLOOKUPS: Found integral MEXPT.
INTEGRALLOOKUPS: Found integral MEXPT.
(%o1)                   log(1 - t) log(t) + li (1 - t)
                                              2
(%i2) diff(%,t);
                                  log(1 - t)
(%o2)                             ──────────
                                      t
~~~


---

**[bugs:#4581] integrate ((log (1 - t)) / t, t, 0, x) is wrong**

**Status:** open
**Group:** None
**Labels:** integrate 
**Created:** Sun Jul 20, 2025 05:19 PM UTC by Raymond Toy
**Last Updated:** Sun Jul 20, 2025 09:13 PM UTC
**Owner:** nobody


In current maxima, we have
```maxima
(%i1) display2d:false;

(%o1) false
(%i2) integrate ((log (1 - t)) / t, t, 0, x);
Is x positive, negative or zero?

pos;
Is x-1 positive, negative or zero?

neg;

(%o2) (log(1-x)*(2*log(x)-2*log(x-1))+log(1-x)^2-2*li[2](-(1/(x-1))))/2+%pi^2
                                                                        /6
```
According to the user manual, this used to return `-li[2](x)`.

If you differentiate `%o2` and manipulate it, you get the negative of the integrand.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4585 Taylor polynomials involving tangent & a quotient

[Barton W. <wil...@us...>]

---

**[bugs:#4585] Taylor polynomials involving tangent & a quotient**

**Status:** open
**Group:** None
**Labels:** taylor 
**Created:** Wed Jul 23, 2025 08:18 PM UTC by Barton Willis
**Last Updated:** Wed Jul 23, 2025 08:18 PM UTC
**Owner:** nobody


This is, I think, OK
~~~
(%i34) taylor(tan(x)/(x-a),x,a,2);

(%o34) tan(a)/(x-a)+(tan(a)^2+1)+(tan(a)^3+tan(a))*(x-a)
                   +((3*tan(a)^4+4*tan(a)^2+1)*(x-a)^2)/3
~~~
But this is wrong: it should be more-or-less `%o34` with `%pi + 1` substituted for `a`:
~~~
(%i36) taylor(tan(x)/(x-%pi-1),x,%pi+1,2);
(%o36) 1+(x-%pi-1)^2/3
~~~


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4584 integrate(1/(x^4+x+1), x, minf, inf)

[Barton W. <wil...@us...>]

---

**[bugs:#4584] integrate(1/(x^4+x+1),x,minf,inf)**

**Status:** open
**Group:** None
**Labels:** definite integral 
**Created:** Tue Jul 22, 2025 06:25 PM UTC by Barton Willis
**Last Updated:** Tue Jul 22, 2025 06:25 PM UTC
**Owner:** nobody


The answer is terribly messy, but this should not happen:
~~~
(%i2)	integrate(1/(x^4+x+1),x,minf,inf);
Message from the stdout of Maxima: Welcome to LDB, a low-level debugger for the Lisp runtime environment.
~~~


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4577 Update CSS for examples and nav bar

[Raymond T. <rt...@us...>]

- **status**: open --> closed
- **Comment**:

Fixed in commit [87ecbc]



---

**[bugs:#4577] Update CSS for examples and nav bar**

**Status:** closed
**Group:** None
**Labels:** Documentation html maxima manual 
**Created:** Wed Jul 16, 2025 04:24 PM UTC by Raymond Toy
**Last Updated:** Wed Jul 16, 2025 05:00 PM UTC
**Owner:** Raymond Toy


When viewed with a narrow window in a browser, the background of examples only extends to the width of the page even if the actual example extends way beyond it.  This looks pretty ugly.  To make it look nicer, a scrollbar for the example could be added so that this doesn't happen and the user can use it to scroll the content instead having it run past the right edge.

Also, the Emacs manual puts a background color on the navigation bar.  It would be nice if we did this too so that the navbar is visually separated from the actual text of the page.


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4583 cdf_general_finite_discrete needs more documentation

[Raymond T. <rt...@us...>]

- **status**: open --> closed
- **assigned_to**: Raymond Toy
- **Comment**:

Fixed in commit [aa96b9]



---

**[bugs:#4583] cdf_general_finite_discrete needs more documentation**

**Status:** closed
**Group:** None
**Labels:** Documentation 
**Created:** Sun Jul 20, 2025 05:28 PM UTC by Raymond Toy
**Last Updated:** Sun Jul 20, 2025 05:28 PM UTC
**Owner:** Raymond Toy


The documentation for general finite discrete random variables doesn't describe the expected values for the integers. I had assumed it was 0, 1, 2, ..., n-1, but it appears to be 1, 2, ..., n.  We should make that more explicit.

This is easy to see from the value of `pdf_general_finite_discrete(2,[1/7,4/7,2/7])` which returns `4/7`, not `2/7`.  Just need to make this explicit in the manual.


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4581 integrate ((log (1 - t)) / t, t, 0, x) is wrong

[Raymond T. <rt...@us...>]

Some random guesses and tracing shows that `rischint` is doing the right thing, but somewhere along the line after `antideriv` returns, we get the wrong result.

This looks right:
```maxima
(%i32) risch(log(1-t)/t,t);
(%o32)                  log(1 - t) log(t) + li (1 - t)
                                              2
```
The derivative is the integrand.  To find the definite integral, we need  to the limit as `t -> 0`.  This is `%pi^2/6`.  I don't know what happens with the upper limit.


---

**[bugs:#4581] integrate ((log (1 - t)) / t, t, 0, x) is wrong**

**Status:** open
**Group:** None
**Labels:** integrate 
**Created:** Sun Jul 20, 2025 05:19 PM UTC by Raymond Toy
**Last Updated:** Sun Jul 20, 2025 06:40 PM UTC
**Owner:** nobody


In current maxima, we have
```maxima
(%i1) display2d:false;

(%o1) false
(%i2) integrate ((log (1 - t)) / t, t, 0, x);
Is x positive, negative or zero?

pos;
Is x-1 positive, negative or zero?

neg;

(%o2) (log(1-x)*(2*log(x)-2*log(x-1))+log(1-x)^2-2*li[2](-(1/(x-1))))/2+%pi^2
                                                                        /6
```
According to the user manual, this used to return `-li[2](x)`.

If you differentiate `%o2` and manipulate it, you get the negative of the integrand.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4581 integrate ((log (1 - t)) / t, t, 0, x) is wrong

[Barton W. <wil...@us...>]

I've checked this against some other standard--I think it is correct:
~~~
(%i22) integrate ((log (1 - t)) / t, t, 0, 1/2);

(%o22) (2*log(2)^2-%pi^2)/4+%pi^2/6
(%i23) expand(float(%));

(%o23) -0.5822405264650126
~~~
This is wrong:
~~~

(%i24) integrate ((log (1 - t)) / t, t, 0, x);
Is x positive, negative or zero?

p;
Is x-1 positive, negative or zero?

n;
(%o24) (log(1-x)*(2*log(x)-2*log(x-1))+log(1-x)^2-2*li[2](-(1/(x-1))))/2+%pi^2/6
(%i25) expand(float(subst(x=1/2,%)));

(%o25) 4.355172180607204*%i-0.5822405264650126
~~~


---

**[bugs:#4581] integrate ((log (1 - t)) / t, t, 0, x) is wrong**

**Status:** open
**Group:** None
**Labels:** integrate 
**Created:** Sun Jul 20, 2025 05:19 PM UTC by Raymond Toy
**Last Updated:** Sun Jul 20, 2025 05:22 PM UTC
**Owner:** nobody


In current maxima, we have
```maxima
(%i1) display2d:false;

(%o1) false
(%i2) integrate ((log (1 - t)) / t, t, 0, x);
Is x positive, negative or zero?

pos;
Is x-1 positive, negative or zero?

neg;

(%o2) (log(1-x)*(2*log(x)-2*log(x-1))+log(1-x)^2-2*li[2](-(1/(x-1))))/2+%pi^2
                                                                        /6
```
According to the user manual, this used to return `-li[2](x)`.

If you differentiate `%o2` and manipulate it, you get the negative of the integrand.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4583 cdf_general_finite_discrete needs more documentation

[Raymond T. <rt...@us...>]

---

**[bugs:#4583] cdf_general_finite_discrete needs more documentation**

**Status:** open
**Group:** None
**Labels:** Documentation 
**Created:** Sun Jul 20, 2025 05:28 PM UTC by Raymond Toy
**Last Updated:** Sun Jul 20, 2025 05:28 PM UTC
**Owner:** nobody


The documentation for general finite discrete random variables doesn't describe the expected values for the integers. I had assumed it was 0, 1, 2, ..., n-1, but it appears to be 1, 2, ..., n.  We should make that more explicit.

This is easy to see from the value of `pdf_general_finite_discrete(2,[1/7,4/7,2/7])` which returns `4/7`, not `2/7`.  Just need to make this explicit in the manual.


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4582 residue(sin(1/(z-1)), z, 1);

[Barton W. <wil...@us...>]

---

**[bugs:#4582] residue(sin(1/(z-1)), z, 1);**

**Status:** open
**Group:** None
**Labels:** residue 
**Created:** Sun Jul 20, 2025 05:25 PM UTC by Barton Willis
**Last Updated:** Sun Jul 20, 2025 05:25 PM UTC
**Owner:** nobody


The residue should be 1:
~~~
(%i4)	residue(sin(1/(z-1)), z, 1);
residue: taylor failed.
 -- an error. To debug this try: debugmode(true);
~~~
I think that a cure might be to use the alternative taylor call:
~~~
(%i3)	taylor (sin(1/(z-1)), [z, 1, 5, asymp]);
(%o3)/T/	1/(z-1)-1/(6*(z-1)^3)+1/(120*(z-1)^5)+...
~~~


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4581 integrate ((log (1 - t)) / t, t, 0, x) is wrong

[Raymond T. <rt...@us...>]

Some numerical evaluation shows that the new result has the correct realpart, but the imaginary part is definitely not close to 0.


---

**[bugs:#4581] integrate ((log (1 - t)) / t, t, 0, x) is wrong**

**Status:** open
**Group:** None
**Labels:** integrate 
**Created:** Sun Jul 20, 2025 05:19 PM UTC by Raymond Toy
**Last Updated:** Sun Jul 20, 2025 05:19 PM UTC
**Owner:** nobody


In current maxima, we have
```maxima
(%i1) display2d:false;

(%o1) false
(%i2) integrate ((log (1 - t)) / t, t, 0, x);
Is x positive, negative or zero?

pos;
Is x-1 positive, negative or zero?

neg;

(%o2) (log(1-x)*(2*log(x)-2*log(x-1))+log(1-x)^2-2*li[2](-(1/(x-1))))/2+%pi^2
                                                                        /6
```
According to the user manual, this used to return `-li[2](x)`.

If you differentiate `%o2` and manipulate it, you get the negative of the integrand.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4581 integrate ((log (1 - t)) / t, t, 0, x) is wrong

[Raymond T. <rt...@us...>]

---

**[bugs:#4581] integrate ((log (1 - t)) / t, t, 0, x) is wrong**

**Status:** open
**Group:** None
**Labels:** integrate 
**Created:** Sun Jul 20, 2025 05:19 PM UTC by Raymond Toy
**Last Updated:** Sun Jul 20, 2025 05:19 PM UTC
**Owner:** nobody


In current maxima, we have
```maxima
(%i1) display2d:false;

(%o1) false
(%i2) integrate ((log (1 - t)) / t, t, 0, x);
Is x positive, negative or zero?

pos;
Is x-1 positive, negative or zero?

neg;

(%o2) (log(1-x)*(2*log(x)-2*log(x-1))+log(1-x)^2-2*li[2](-(1/(x-1))))/2+%pi^2
                                                                        /6
```
According to the user manual, this used to return `-li[2](x)`.

If you differentiate `%o2` and manipulate it, you get the negative of the integrand.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.