Matlab: usage of split dataset spliteachlabel()

splitEachLabel yes ImageDatastore An object function of object ……
Of course , It doesn't matter
It's important to know splitEachLabel The input is one ImageDatastore object

The usage function ：
[imds1,imds2] = splitEachLabel(imds,p)
[imds1,…,imdsM] = splitEachLabel(imds,p1,…,pN)

[imds1,imds2] = splitEachLabel(imds,p) Can be imds The image file in is split into two new data stores ,imds1 and imds2. New data storage imds1 Include the front of each label p The file of ,imds2 The remaining files that contain each tag .p It can be 0 To 1 Number between , Indicate that each tag is assigned to imds1 Percentage of files ; It can also be an integer , Indicate that each tag is assigned to imds1 Absolute number of files .

[imds1,…,imdsM] = splitEachLabel(imds,p1,…,pN) Split the data store into N+1 A new data store . The first new data store imds1 Include the front of each label p1 The file of , Next new data store imds2 Include the following p2 The file of , And so on . If p1,…,pN Represents the number of documents , Then their sum cannot exceed the original data storage imds The minimum number of files in the label .