Inifinity symbol and trigonometric functions #681
-
|
Hi there! How can I get the infinity value symbol without the plus/minus (as shown in line 270 of OutputWriter.cs)??? 
Since I had no success in my previous attempts, I decided to try some trigonometric functions. Then I found out that the results from such functions may need some tweaks, no? Maybe forcing some zero/infinity values for very known results? Or maybe forcing some rounding? 
|
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 5 replies
-
|
Hi! Calcpad uses the standard 64 bit floating point math functions as they are in the CPU and exposed via the Math class in C#. You will get the same results in any other language as C, C++, C#, Java, etc. The way they are computed in floating point environment is specified by the IEEE 754 standard. The best way is to leave them like this. Adding some artificial corrections/rounding to the values can be dangerous. But If you want to see them rounded, you can apply the respective number formatting. The things are different for symbolic CAS and arbitrary precision calculation systems. |
Beta Was this translation helpful? Give feedback.
-
|
To your question about infinity: Why is that? For real numbers, there is no "absolute" infinity. It is either on the left (-∞) or right (+∞) side of the number line although the "+" sign is usually omitted. In the complex domain, things are different. There is only one infinity ∞ that is defined by mapping the complex plane to a Riemann sphere and thus extending the complex domain: The complex infinity does not have any signs. Things are also interesting about division by zero in real domain. Mathematically it is not allowed, but it is defined for double precision arithmetic by the IEEE 754 standard as follows: https://en.wikipedia.org/wiki/IEEE_754 |
Beta Was this translation helpful? Give feedback.
-
|
Hi, @Proektsoftbg! Thanks for the explanation. I was aware of all those Math, IEEE standard and floating point representations. To be honest, I get similar results in my Fortran programs (and it's the same with other programming languages). I just thought we could "force" the correct answer to known "exact" values. I was thinking about something like the "zero small elem." option for vectors and matrices. But I just test it and I found out it's just a "cosmetic" output stuff anyway, since it didn't really force the element to be zero... I will just keep using number formatting when necessary. About the infinity representation... Well... Complex mode isn't suitable for me at all. As you pointed out, I just wanted to omit the plus sign before the positive infinity symbol (and I tried to use |
Beta Was this translation helpful? Give feedback.
-
|
It is interesting why do you want to omit the + for infinity time? It is also valid there. At least, it would quite acceptable for me. |
Beta Was this translation helpful? Give feedback.
-
|
Yeah, @Proektsoftbg, it's perfectly fine and I was already using it like this... 
Since I'm using Calcpad to share some problems solutions with my students, I want the calculations to be exactly like they should do "by hand". That's why I wanted to omit the plus sign. |
Beta Was this translation helpful? Give feedback.
-
|
Aha! OK, I see now. Thank you. |
Beta Was this translation helpful? Give feedback.
Hi! Calcpad uses the standard 64 bit floating point math functions as they are in the CPU and exposed via the Math class in C#. You will get the same results in any other language as C, C++, C#, Java, etc. The way they are computed in floating point environment is specified by the IEEE 754 standard.
The best way is to leave them like this. Adding some artificial corrections/rounding to the values can be dangerous. But If you want to see them rounded, you can apply the respective number formatting. The things are different for symbolic CAS and arbitrary precision calculation systems.