How do I convert a numerical vector into a comma-delimited string?
313 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
Given a numerical vector, how do I convert it into a comma-delimited string?
For example, if
n = [12345 6789 10234 3452]
convert n to
12345,6789,10234,3452
0 commentaires
Réponse acceptée
Plus de réponses (6)
Michael
le 15 Juin 2018
Have another one-liner:
regexprep(num2str(n),'\s+',',')
If uninitiated in the secrets of (Matlab's) regular expressions: \s stands for any whitespace and the + means 1 or more times. This is needed because the amount of space between the numbers output by num2str varies depending on the length of those numbers.
0 commentaires
Jan Siegmund
le 30 Juin 2020
Here is another oneliner:
strjoin(cellstr(num2str(n')),',')
Testing all four types with a script
n = randi(100,[1 100]);
tic;
s1 = sprintf('%.0f,' , n);
s1 = s1(1:end-1);
toc
tic;
s2 = strjoin(arrayfun(@(x) num2str(x),n,'UniformOutput',false),',');
toc
tic;
s3 = regexprep(num2str(n),'\s+',',');
toc
tic;
s4 = strjoin(cellstr(num2str(n')),',');
toc
deliveres:
Elapsed time is 0.000248 seconds.
Elapsed time is 0.005549 seconds.
Elapsed time is 0.000746 seconds.
Elapsed time is 0.003352 seconds.
The accepted answer is the fastest.
0 commentaires
Dyuman Joshi
le 18 Août 2023
(Quite) A simple approach that has not been mentioned yet -
n = [12345 6789 10234 3452];
out = strjoin(string(n),",")
3 commentaires
Bruno Luong
le 18 Août 2023
Modifié(e) : Bruno Luong
le 18 Août 2023
The first method is not optimal with format '%.0f', for integer better using '%i' (method 6)
n = randi([1e3 1e4],[1 1000]);
t = zeros(1,6);
t(1) = timeit(@() csnum1(n), 1);
t(2) = timeit(@() csnum2(n), 1);
t(3) = timeit(@() csnum3(n), 1);
t(4) = timeit(@() csnum4(n), 1);
t(5) = timeit(@() csnum5(n), 1);
t(6) = timeit(@() csnum6(n), 1);
tms = t*1e3
function s1 = csnum1(n)
s1 = sprintf('%.0f,' , n);
s1 = s1(1:end-1);
end
function s2 = csnum2(n)
s2 = strjoin(arrayfun(@(x) num2str(x),n,'UniformOutput',false),',');
end
function s3 = csnum3(n)
s3 = regexprep(num2str(n),'\s+',',');
end
function s4 = csnum4(n)
s4 = strjoin(cellstr(num2str(n')),',');
end
function s5 = csnum5(n)
s5 = strjoin(string(n),',');
end
function s = csnum6(n)
s = sprintf('%i,' , n);
s = s(1:end-1);
end
Dyuman Joshi
le 19 Août 2023
Even after the change, Method 6 is (marginally) slower than Method 5 (which is the fastest).
RST
le 3 Août 2023
We can also use compose() for a formatted one-liner:
a = rand(5,1);
csva = strjoin(compose('%.3f', a), ', ')
and also for 2D arrays:
b = rand(3,4);
csvb = strjoin(join(compose( '%.3f', b), ', '), newline())
0 commentaires
Gianfranco Di Domenico
le 18 Août 2023
Modifié(e) : Gianfranco Di Domenico
le 18 Août 2023
EDIT: The backspace character is handled differently according to the final "usage" of the string. For instance, it works perfectly when displaying the string in the terminal; however, printing the string to a file does not remove the comma. Moreover, the byte array corresponding to the string retains both the comma and the backspace character.
More elegant, one liner:
n = [12345 6789 10234 3452]
n_str = sprintf("%s\b", sprintf("%i,", n))
The logic:
- The nested sprintf creates a string of comma-separated values, but has a trailing comma
- The top-level sprintf prints the generated strings and the \b character (backspace) removes the trailing comma
3 commentaires
Gianfranco Di Domenico
le 18 Août 2023
I just noticed this behavior.
I will edit the original answer because it seems that the behavior depends on the final usage of the string. For instance, fprintf-ing to a file does NOT work! The comma is retained.
Bruno Luong
le 18 Août 2023
Modifié(e) : Bruno Luong
le 18 Août 2023
Example
n = [12345 6789 10234 3452]
n_str = sprintf("%s\b", sprintf("%i,", n))
str2num(n_str) % oh oh ...
str2num(extractBefore(n_str,strlength(n_str)-1))
This might create some frustration and pain of debugging
Voir également
Catégories
En savoir plus sur Performance and Memory dans Help Center et File Exchange
Produits
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!