Re: Efficient sort function allowing own test and swap function as parameter
On 10/6/20 3:18 PM, Alaindevos wrote:
I have a large table consisting of two columns.One with words.Another
with frequencies. I want to sort them efficiently according to the names
or frequency.
For this I need an efficient sort function where I can plugin my proper
test of order, and proper swap. Currently I do it using an own written
bubble sort that doesn't scale well.
I had fun writing the following program. Note how makeIndex allows
visiting elements in sorted order without actually sorting them.
import std.random;
import std.range;
import std.algorithm;
import std.conv;
import std.stdio;
struct S {
string word;
size_t frequency;
}
bool byWord(S a, S b) {
return a.word < b.word;
}
bool byFrequency(S a, S b) {
return a.frequency < b.frequency;
}
auto dump(R)(string title, R range) {
writefln!"\n%s:\n%(%s\n%)"(title, range);
}
// A test function that makes an S
S makeS() {
string makeWord() {
static letters = iota('a', 'z' + 1).map!(to!dchar).array;
return letters.randomSample(4).to!string; // Four-letter words! :p
}
size_t makeFrequency() {
return uniform(0, 100);
}
return S(makeWord(), makeFrequency());
}
// A test function that makes some S'es
S[] makeSs() {
return 10.iota.map!(i => makeS()).array;
}
void main() {
auto ss = makeSs();
dump("Unsorted", ss);
auto byWordIndexes = new size_t[ss.length];
ss.makeIndex!byWord(byWordIndexes);
dump("Still unsorted but visited by word order",
byWordIndexes.map!(i => ss[i]));
auto byFrequencyIndexes = new size_t[ss.length];
ss.makeIndex!byFrequency(byFrequencyIndexes);
dump("Still unsorted but visited by frequency order",
byFrequencyIndexes.map!(i => ss[i]));
ss.sort!byWord();
dump("Actually sorted by words", ss);
ss.sort!byFrequency();
dump("Actually sorted by frequencies", ss);
}
Sample output:
Unsorted:
S("bfmp", 78)
S("imsx", 17)
S("kmwy", 60)
S("klpw", 92)
S("hnrt", 24)
S("aivz", 29)
S("prst", 24)
S("cdlm", 86)
S("alvz", 13)
S("mnxz", 52)
Still unsorted but visited by word order:
S("aivz", 29)
S("alvz", 13)
S("bfmp", 78)
S("cdlm", 86)
S("hnrt", 24)
S("imsx", 17)
S("klpw", 92)
S("kmwy", 60)
S("mnxz", 52)
S("prst", 24)
Still unsorted but visited by frequency order:
S("alvz", 13)
S("imsx", 17)
S("hnrt", 24)
S("prst", 24)
S("aivz", 29)
S("mnxz", 52)
S("kmwy", 60)
S("bfmp", 78)
S("cdlm", 86)
S("klpw", 92)
Actually sorted by words:
S("aivz", 29)
S("alvz", 13)
S("bfmp", 78)
S("cdlm", 86)
S("hnrt", 24)
S("imsx", 17)
S("klpw", 92)
S("kmwy", 60)
S("mnxz", 52)
S("prst", 24)
Actually sorted by frequencies:
S("alvz", 13)
S("imsx", 17)
S("hnrt", 24)
S("prst", 24)
S("aivz", 29)
S("mnxz", 52)
S("kmwy", 60)
S("bfmp", 78)
S("cdlm", 86)
S("klpw", 92)
Ali
Re: Efficient sort function allowing own test and swap function as parameter
On Tue, Oct 06, 2020 at 10:18:39PM +, Alaindevos via Digitalmars-d-learn wrote: > I have a large table consisting of two columns.One with words.Another > with frequencies. I want to sort them efficiently according to the > names or frequency. > For this I need an efficient sort function where I can plugin my > proper test of order, and proper swap. Currently I do it using an own > written bubble sort that doesn't scale well. Why don't you use std.algorithm.sort? That's what the standard library is for. T -- Being able to learn is a great learning; being able to unlearn is a greater learning.
Efficient sort function allowing own test and swap function as parameter
I have a large table consisting of two columns.One with words.Another with frequencies. I want to sort them efficiently according to the names or frequency. For this I need an efficient sort function where I can plugin my proper test of order, and proper swap. Currently I do it using an own written bubble sort that doesn't scale well.
Re: It is possible to substract 5 from 3 unsigned integer
On Tuesday, 6 October 2020 at 18:24:14 UTC, Alaindevos wrote: There are two subtractions possible. A machine-one which can be architecture dependent, does not have the same results on all computers, and behaves like a modulus in mathematics. A logical one. For the last one higher classes might be needed. The (signed and unsigned) integer wrap around is well-defined in D, and is architecture independent. That means for example that if a specific architecture does not have a subtract instruction that wraps around according to what D specifies, then the compiler will generate the necessary instructions such that still int.min - 1 == int.max. That's similar to how integer multiplication 'just works' on architectures that do not have a multiply instruction. -Johan
Re: It is possible to substract 5 from 3 unsigned integer
On Tuesday, 6 October 2020 at 18:24:14 UTC, Alaindevos wrote: There are two subtractions possible. A machine-one which can be architecture dependent, does not have the same results on all computers, and behaves like a modulus in mathematics. A logical one. For the last one higher classes might be needed. Hello, You may try https://dlang.org/phobos/std_experimental_checkedint.html in this case. Regards,
Re: It is possible to substract 5 from 3 unsigned integer
There are two subtractions possible. A machine-one which can be architecture dependent, does not have the same results on all computers, and behaves like a modulus in mathematics. A logical one. For the last one higher classes might be needed.
Re: Link Time Optimization Bitcode File Format
On Tuesday, 6 October 2020 at 16:46:28 UTC, Severin Teona wrote: Also, when I try to link the druntime with the application I want to write on the microcontroller, there are some link errors due to the file format. This happens when you link manually, not through LDC. When running LDC with `-flto=full -O -v ...`, you'll see that it invokes gcc with extra flags, something like -Wl,-plugin,/dlang/ldc-1.23.0/lib/LLVMgold-ldc.so -Wl,-plugin-opt=mcpu=x86-64 -Wl,-plugin-opt=O3 -Wl,-plugin-opt=-function-sections -Wl,-plugin-opt=-data-sections so that the linker gets to know how to deal with bitcode objects via the LLVM plugin. See http://johanengelen.github.io/ldc/2016/11/10/Link-Time-Optimization-LDC.html for more infos about LTO.
Re: Link Time Optimization Bitcode File Format
On Tuesday, 6 October 2020 at 16:46:28 UTC, Severin Teona wrote: Hi all, I am trying to build the druntime with the 'ldc-build-runtime' tool for microcontrollers (using the arm-none-eabi-gcc compiler) and therefore the size of the druntime should be as little as possible. One solution I had was to use Link Time Optimization (LTO) to reduce the size. The problem now is the fact that when I compile the druntime with -flto=full or -flto=thin (as arguments for LDC2), the resulted object files (and also a big part of the runtime as it is a static library) have a different file format - LLVM IR bitcode - than I need, which is ELF 32-bit. Also, when I try to link the druntime with the application I want to write on the microcontroller, there are some link errors due to the file format. I also tried using a different archiver - llvm-ar - but I had no luck. Could you give me some advice about how should I fix this? Thank you! I have run into the same problem when using GNU ld. The problem is that my version of GNU ld, version 2.30.0.20180329 (which is ancient) cannot deal with the object file format when LTO is enabled. One way is to try llvm-ld and see if that one can handle it.
Link Time Optimization Bitcode File Format
Hi all, I am trying to build the druntime with the 'ldc-build-runtime' tool for microcontrollers (using the arm-none-eabi-gcc compiler) and therefore the size of the druntime should be as little as possible. One solution I had was to use Link Time Optimization (LTO) to reduce the size. The problem now is the fact that when I compile the druntime with -flto=full or -flto=thin (as arguments for LDC2), the resulted object files (and also a big part of the runtime as it is a static library) have a different file format - LLVM IR bitcode - than I need, which is ELF 32-bit. Also, when I try to link the druntime with the application I want to write on the microcontroller, there are some link errors due to the file format. I also tried using a different archiver - llvm-ar - but I had no luck. Could you give me some advice about how should I fix this? Thank you!
Re: It is possible to substract 5 from 3 unsigned integer
On 10/6/20 5:24 AM, Alaindevos wrote:
Is that the expected behavior of the programmer?
Opinions can differ. Feel free to elaborate.
The following is even more "expected". ;) Subtract zero from -1 and you
get size_t.max.
void main() {
int[] arr;
int i = -1;
auto u = (i - arr.length);// -1 - 0
assert(u == size_t.max); // the surprise
static assert (is (typeof(u) == size_t)); // the reason
}
Ali
Re: It is possible to substract 5 from 3 unsigned integer
On Tuesday, 6 October 2020 at 12:24:56 UTC, Alaindevos wrote: Is that the expected behavior of the programmer? Opinions can differ. Feel free to elaborate. It's expected behavior: "If both operands are of integral types and an overflow or underflow occurs in the computation, wrapping will happen. For example, uint.max + 1 == uint.min, uint.min - 1 == uint.max, int.max + 1 == int.min, and int.min - 1 == int.max." https://dlang.org/spec/expression.html#add_expressions
Re: It is possible to substract 5 from 3 unsigned integer
On Tuesday, 6 October 2020 at 12:24:56 UTC, Alaindevos wrote: Is that the expected behavior of the programmer? Opinions can differ. Feel free to elaborate. Elaborate please. Are you really asking if one can do subtraction in D.
Re: It is possible to substract 5 from 3 unsigned integer
On Tuesday, 6 October 2020 at 12:24:56 UTC, Alaindevos wrote: Is that the expected behavior of the programmer? Opinions can differ. Feel free to elaborate. E.g. length of a string unsigned long.
It is possible to substract 5 from 3 unsigned integer
Is that the expected behavior of the programmer? Opinions can differ. Feel free to elaborate.
Re: vibe.d / experience / feedback
On Thursday, 1 October 2020 at 06:32:23 UTC, Robert M. Münch wrote: Hi, we are currently using vibe.d for a prototype and I want to post some experiences. I know one shouldn't only address the problems but provide some solutions. However, our current use-case is that we want to get a job done, and not creating a web-framework. ... So, overall... compared to other web-frameworks we know, there are many pitfalls which just cost a lot of time... We think it's a good base, but there needs to be much less magic, much better docs, much less complexity. Hi Robert, I found myself in a similar situation recently, and I can't help but ask you: What technology do you use regularly? What drives/draws you to try dlang/vibe.d? Do you have other alternatives to dlang/vibe.d for your project? In my case we usually work in Node+js/ts (previously Scala+Play) and I wanted to jump to something really performant for a new project without losing code expressiveness and development speed. Dlang seemed a good alternative (I like it much more than Go or Rust).
Re: vibe.d / experience / feedback
On Thursday, 1 October 2020 at 06:32:23 UTC, Robert M. Münch wrote: Hi, we are currently using vibe.d for a prototype and I want to post some experiences. I know one shouldn't only address the problems but provide some solutions. However, our current use-case is that we want to get a job done, and not creating a web-framework. 1. For many things the docs are missing, or inconsistent, or wrong. So, it's pretty hard to make your way through it. Or you just start reading the code and reverse engineer vibe. => We are waisting a lot of time. 2. There are many inconsistencies / half-done implementations in vibe. Again, this makes it hard to find out what works, what doesn't or which parts behave differently. => Costs a lot of time. 3. Vibe can't handle GET & POST on the same URL... and the bad thing is, the later route will just be ignored, so you don't get any hint or crash. The docs don't mention this clearly. => We waisted a lot of time to find out, that this doesn't work. 4. Vide uses an own JSON type, not the standard one. We don't understand why, this just makes things much more complicated and one has to mess around with this. 5. One can't access the raw HTTP request body, things must be go through Vibe's JSON parser. To get access to the raw body, a lot of workarounds are necessary. So, overall... compared to other web-frameworks we know, there are many pitfalls which just cost a lot of time... We think it's a good base, but there needs to be much less magic, much better docs, much less complexity. I concur with 1 and 2. We waste lot of time trying to understand vibe code instead of working on the business logic. Never had issues with 3. I've found vibe JSON to be a relative of nlohmann::json https://github.com/nlohmann/json Similar API, seamless code reusability (copy paste). I never really cared about std.json as the usability is not good. 5. is definitely possible with readAll. Never had issues with this. Just dig more into the code ☺️
