| >
Specific commands for datasets |
|
display | writes in the command window the name, numbers of objects, features and classes as well as class sizes. |
|
disp | shows the contents of the data field. disp(A) is the same as disp(getdata(A)). |
|
double | converts the dataset into doubles by returning the data field. double(A) is the same as getdata(A). |
|
+A | is a short for double(A). |
|
show | displays images that are stored in A, see datasets examples. |
|
find | find(A) is identical to find(+A) and returns the indices of the non-zero values of the data in A. |
|
findfeatlab | returns the indices(column numbers) of the given features. |
|
findlabels | returns indices of all objects with given labels. |
|
findnlab | returns the indices of all objects from given class numbers. |
|
getclassi | finds the class numbers for given class names. |
|
classnames | retrieves the names of the classes, also called the label list. |
|
classsizes | retrieves the sizes of the classes, i.e. the number of objects per class. |
|
renumlab | find a label list from a set of labels, or merges two label lists. |
|
matchlablist | matches label lists. |
|
misval | fix missing values. |
|
labcmp | finds differences between two sets of labels. |
|
nlabcmp | counts the differences between two sets of labels. |
|
remclass | removes a class from the dataset. |
|
seldat | selects classes, features or objects from a dataset. |
|
setdat | changes the data field of a dataset with new data. This command is very similar to setdata, but adds an option to change the feature labels simultaneously in relation with classifier outputs. |
|
isdataset | tests variable on dataset type. |
|
isvaldset | tests on dataset validity for some problem: number of classes, numbers of objects. |
|
im2obj | converts a set of images to a dataset: images become objects, see image handling. |
|
im2feat | converts a set of images to a dataset: images become features, see image handling. |
|
|
