在MATLAB中按值排序数据

Question

我有一个包含各种值的数据表。它包括以下几列:

| Value 1 | Value 2 |
|---------|---------|
|0        |235      |
|1        |123      |
|1        |309      |
|1        |540      |
|2        |34       |
|3        |123      |
|3        |959      |
|3        |3939     |

我想有一个表，其中Value 2取决于Value 1的每个值。它看起来像这样:

| Value 1 | Value 2        |
| --------|----------------|
| 0       |[235]           |
|1        |[123, 309, 540] |
|2        |[34]            |
|3        |[123, 959, 3939]|

之类的

我尝试了几种解决方案，例如:

t = table;
for i=1:length(existing_table)  
t(existing_table, end+1) = existing_table.value2(i) % append to case i
end;

--> returns Error using  () 
Right hand side of an assignment into a table must be another table or a cell array.

--------------
t = table;
for i=1:length(existing_table)  
switch existing_table.value2(i) 
case 0
t(0, end+1) = existing_table.value2(i) % append to case 
0
% etc...
end;
end;


--> returns Error using  () 
Right hand side of an assignment into a table must be another table or a cell array.
--------------
t = table;
t.v1 = value1;
t.v2 = unique(value2) % Does not return the kind of table that I would like

Answer 1

这听起来像是accumarray的工作:

% Create sample data
value1 = [0 1 1 1 2 3 3 3].';
value2 = [235 123 309 540 34 123 959 3939].';
T = table(value1,value2);
% Accumarray handles only positive indexes. Get these in idx using unique
[uni,~,idx] = unique(T.value1(:));
% First colum is just the unique value in the first column of T
value1_out = uni;
% accumarray groups the values corresponding to each index. Then @(x) {x}
% gathers it into a cell array
value2_out = accumarray(idx,T.value2,[],@(x) {x});
Tout = table(value1_out,value2_out);