Had to change unique in levels in constructDataset function so markers order can be preserved --> the predictions from pRoloc weren't good
unique changed order of class prediction so when calling : labels <- levels(.trainSep$theLabels) in pRolocClassification the predictions labels weren't well assigned.
Hello,
Had to change unique in levels in constructDataset function so markers order can be preserved --> the predictions from pRoloc weren't good unique changed order of class prediction so when calling : labels <- levels(.trainSep$theLabels) in pRolocClassification the predictions labels weren't well assigned.