I would like to generate random data using R or RStudio for a distribution with function as below from scratch. How do I develop the codes. I am new to this type of models.
Related
Usually, i'm using "lcmm" package in R - in order to perform growth mixture modeling (GMM).
However, i want to create a classification or groups based on two variables rather than one. I've read that this kind of modeling called "parallel growth mixture modeling". Is anyone knows how to perform such an analysis using R script?
I clustered some data I have using FactoMineR::HCPC in R and now I am trying to cluster new data using the model I have trained. However I can not find a function that will give me such prediction. So far I've been able to see that I can use predict.PCA with my new data set and the previous PCA I implemented on the trainig data set and use that result on the function HCPC. I know that for kmenas there exist a predict which will take as input the trained model and the new data. Does anybody know if there is an equivalent for HCPC?
How to save caret trained models so that it can be used later for building ensemble models in RStudio?
Is it currently possible to create a custom R clustering model, where you can define your own clustering model? Because AzureML does not let you connect Customer R Model with Train Clustering Model.
This is a critical limitation of AzureML when it comes to clustering.
Note: I know that you can create it in Execute R Script, but I want to be able to save the model so when new test data is inputted, I would assign it to the respective clusters.
I am using the kohonen library in R to train a self-organizing map using some data. I split the data set as 60/40 for training/testing purposes. How can I run the trained model with testing data? I can't find a function in the documentation to do this. How can I review how off the test from the training data is? I am using R 3.3.2 and RStudio 1.0.44.