Expon bug in wp34s

[Andrew Nikitin]

Quote:

---

1 [sto] J
0 [sto] K
1 [chs] Expon

---

returns -1.71828182846 instead of correct 0.

I tried to correct it but failed to find code for Expon. I found the distributions.wp34s in xrom folder, but Expon is not there. Where is it?

[Marcus von Cube, Germany]

Look in xrom/distributions/exponential.wp34s.

[Andrew Nikitin]

Here is a patch (I kind of tested it by building, flashing and trying).

Index: trunk/xrom/distributions/exponential.wp34s

===================================================================

--- trunk/xrom/distributions/exponential.wp34s	(revision 3429)

+++ trunk/xrom/distributions/exponential.wp34s	(working copy)

@@ -55,6 +55,8 @@

 			SPEC?

 				JMP cdf_infinite

 			[times]

+                        x<0?

+				JMP ret_0

 			+/-

 			e[^x]-1

 			+/-

@@ -70,7 +72,6 @@

 			xOUT xOUT_NORMAL

 expon_param::		RCL J

-			GSB check_special_pos

-			RTN

+			JMP check_special_pos

 

[Andrew Nikitin]

Since I was there, I think that similar bug in Expon_u should also be fixed.
I tried this one in console emulator.

Index: trunk/xrom/distributions/exponential.wp34s

===================================================================

--- trunk/xrom/distributions/exponential.wp34s	(revision 3429)

+++ trunk/xrom/distributions/exponential.wp34s	(working copy)

@@ -43,6 +43,8 @@

 			SPEC?

 				JMP cdf_infinite

 			[times]

+                        x<0?

+				JMP ret_1

 			+/-

 			e[^x]

 			xOUT xOUT_NORMAL

@@ -55,6 +57,8 @@

 			SPEC?

 				JMP cdf_infinite

 			[times]

+                        x<0?

+				JMP ret_0

 			+/-

 			e[^x]-1

 			+/-

@@ -70,7 +74,6 @@

 			xOUT xOUT_NORMAL

 expon_param::		RCL J

-			GSB check_special_pos

-			RTN

+			JMP check_special_pos

 

Edited: 13 July 2013, 5:01 p.m.

[Paul Dale]

Thanks for this. Patch applied now.

- Pauli

[Andrew Nikitin]

I am not sure what happened, but when I updated from SVN, the Expon_u became this:

		XLBL"CDFU_EXPON"

			xIN MONADIC

			GSB expon_param

			SPEC?

				JMP cdf_infinite

			x<0?

				JMP ret_1

			[times]

			x<0?

				JMP ret_1

			+/-

			e[^x]

			xOUT xOUT_NORMAL

the first instance "x<0? jmp ret_1" (right before [times]) was not in the patch and should not be there.

[Paul Dale]

Hopefully fixed now.

I had to manually apply the patch -- the patch program didn't like me.

Pauli

[Marcus von Cube, Germany]

I've uploaded a new build to SVN. It should fix the RUN vs STOP issue, too.