Hi, you also need to not use subheadings in the task description - they are reserved for Language examples exclusively.
You can get that kind of effect using:
- Not a heading
;What is typed for the above type of bold subheading starts with a semicolon
Although I'd be delighted to be proved wrong, there doesn't appear to be a way to change the tab size on this site which is fixed at 8.
--ChristerNilsson (talk) 15:53, 15 March 2019 (UTC) In Github I just add a file .editorconfig where tab width is stated. I find tabs has a lot of advantages. Smaller files, fewer keystrokes, and each individual might use his/her own tab width. But, this is religion and not even Nim allows tabs. I must use the following directive as the first line in every .nim file: #? replace(sub = "\t", by = " ")
- 1 In general, Rosettacode task names should only have the first word in a title capitalized. (Excluding proper nouns) It isn't a hard rule, and has been broken before, but is the preferred style.
- 2 Is there any particular reason this is named Parallel resistor calculator? Seems like it is doing both parallel and series resistances. I would propose to rename it to "Resistance equivalence calculator" to frontload what the primary point is: "Resistance", more accurately describe what it does, and bring it in line with site naming conventions. (Or even just Resistance calculator ) --Thundergnat (talk) 12:23, 15 March 2019 (UTC)
--ChristerNilsson (talk) 16:02, 15 March 2019 (UTC) I added a Nim solution for a better handling of infix input. Ideally, the source code should not have to be recompiled for every new input. Also, most of the code in Infix and RPN are identical. So, they might be merged.
Input change proposal:
- Infix: 2 + 3
- RPN: 2 3 +
This proposal unfortunately has the implication that operator overloading is no longer needed.
Serial/parallel circuits are a simple case, there is a more general case where you can model the resistor network with a sparse symmetric matrix and solve a linear system to get the resistance between any two points of the network. See Resistor mesh#Maxima for instance. Eoraptor (talk) 23:51, 19 March 2019 (UTC)
The Mesh is very interesting. I haven't figured out how to enter such a beast into a solver using either infix or RPN. The mesh example is very symmetric and easy. How would you propose dealing with a spaghetti mesh? --ChristerNilsson (talk) 01:12, 20 March 2019 (UTC)
Ok, there is a solution, but I think it should stand on it's own feet. Using the picture and labelling the nodes we get: AC6 CH8 CD4 DH8 GH4 DG6 DE10 FG8 DF6 EF10. Then we can use the Mesh code. --ChristerNilsson (talk) 02:23, 20 March 2019 (UTC)
The Mesh example has 180 resistors and 100 nodes. The description might be A1 A2 1|A1 B1 1| .. |J9 J10 1. Then the iteration might start. Should be able to solve my problems as well. --ChristerNilsson (talk) 02:32, 20 March 2019 (UTC)
- A "spaghetti" mesh would still lead to a symmetric matrix: this does not come from the symmetry of the circuit, but from Kirchhoff's laws. If there are N nodes the matrix A has dimensions NxN, and there is a nonzero element A(i,j) for each nodes i,j linked by a resistor. To describe an arbitrary circuit, you need basically to be able to decribe an arbitrary graph: a (usually sparse) matrix, or a list of nodes together with a list of edges, for instance. Eoraptor (talk) 20:17, 21 March 2019 (UTC)