Different notebooks need different enviornments, using a single one caused conflicts
- all R packages glmpca, simlr & scry used singlecell
- Deep Learning autoencoders used - scVAE env
- Zifa & ICA used simple singlecell env
- scvis needs its own env so used scvi-env
Note Tried using separate env for scphere adn grandprix but tensorflow incomaptibility issues , even with older version of gpflow the library would't run Using apple gpus