Simple question regarding bar plot with categorical data
显示 更早的评论
Hello guys,
I am new to matlab, and I'm trying to do a simple bar plot, like this:
x = ["bananas" "apples" "cherries"];
y = [14,12,7];
bar(categorical(x),y);
The problem is that bar() function seems to sort the x-labels in alphabetical order. Is there any way to override this behaviour, so that the x-labels are left in their original order?
采纳的回答
Sean de Wolski
2017-10-6
编辑：Sean de Wolski
2017-10-6
x = categorical(["bananas" "apples" "cherries"]);
x = reordercats(x,{'bananas' 'apples' 'cherries'});
y = [14,12,7];
bar(x,y);
Categoricals can have order associated with them (for the purpose of relational operators).
7 个评论
maho
2017-10-9
Allen
2017-12-21
I am using R2017a and experiencing a similar issue with the sort order of categorical labels in bar charts. However, my issue does appear to be in the bar and not with categorical.
list = {'t1','t2','t3','t4','t5','t6','t7','t8','t9','t10','t11'};
c = categorical(list);
The order of the array matches the order of list. However, when using the following, the categorical axis labels are reordered along with the values in NumVector, such that their new order matches list2 below and is not desired.
bar(c,NumVector)
list2 = {'t1','t10','t11','t2','t3','t4','t5','t6','t7','t8','t9'};
Since the NumVector input values have also been reordered, it is not simply possible to replace the XTickLabel with the values in list as this results in mismatched labeling. Any suggestions on how to get around this issue?
Sean de Wolski
2017-12-22
Allen, ,y default it's sorting the categories. You can pass in the valueset yourself which will enforce order or call reorder like I did above.
list = {'t1','t2','t3','t4','t5','t6','t7','t8','t9','t10','t11'};
c = categorical(list,list);
bar(c,1:numel(c))
Ellyn Gray
2019-5-22
Thanks!
tesarj13
2020-11-23
Even better with coversion to string:
x = categorical(["bananas" "apples" "cherries"]);
x = reordercats(x,string(x));
y = [14,12,7];
bar(x,y);
Starting in R2023b, you can specify x as a string vector (or a cell array of character vectors), and MATLAB preserves the order. For example:
x = ["bananas" "apples" "cherries"];
y = [14,12,7];
bar(x,y)
Henry
2024-10-14
Since you guys are fixing this. Could you also put in a ticket to fix this for the errorbar() plot
更多回答（4 个）
Steven Lord
2017-10-6
If you're using release R2016b or later you could use histogram with a vector of BinCounts instead of using bar.
x = ["bananas" "apples" "cherries"];
C = categorical(x);
y = [14,12,7];
h = histogram('Categories', C, 'BinCounts', y);
I also faced the same issue. There are numerous shortcut techniques available to solve it out. What I used to follow is as below :
x = categorical(["bananas" "apples" "cherries"]);
x = reordercats(x,cellstr(x)');
y = [14,12,7];
bar(x,y);
The problem is not in bar, but in categorical. I can't find in the doc how to preserve order (with unique I know there is a switch to do so). So my suggestion would be to convert it yourself:
[~,ia,~]=unique(x,'stable');
x2=1:length(x);
x2=x(ia);
bar(x2,y)
xticks(x2)
xticklabels(x)%might not work, as this expects a cell stray containing strings
Of course, your labels are very likely to be unique, otherwise bar wil most likely yield an error, so this should work the same in all valid situations:
x2=1:length(x);
bar(x2,y)
xticks(x2)
xticklabels(x)%might not work, as this expects a cell stray containing strings
*** Use histogram and not bar :)
I've been struggling with this until now, so sharing with the community. You can use the histogram simply as a drawing function, not actually counting anything.
Including the vertical bars as well (instead of "barh"), as long labels are typically better displayed horizontally.
Et voila! Good luck. Mike
x = ["bananas" "apples" "cherries"];
y = [14,12,7];
histogram('Categories',x,'BinCounts',y,'orientation','horizontal');
histogram('Categories',x,'BinCounts',y);
类别
在 帮助中心 和 File Exchange 中查找有关 Creating, Deleting, and Querying Graphics Objects 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
