diff --git a/README.md b/README.md
index e7b9b8f4..14c72363 100644
--- a/README.md
+++ b/README.md
@@ -6,7 +6,7 @@ The functions are analogous versions of functions that can be found within [scan
 
 ## Requirements
 
-To run the code in this repository you need a conda environment with rapids and scanpy installed. To use the full functionality of this repo please use `rapids-22.04` or `rapids-22.06`. You also need an Nvidia GPU.
+To run the code in this repository you need a conda environment with rapids and scanpy installed. To use the full functionality of this repo please use `rapids-22.08`. You also need an Nvidia GPU.
 ```
 conda create -n rapids_singelcell -f conda/rapids_singecell.yml
 conda activate rapids_singelcell
@@ -14,11 +14,13 @@ ipython kernel install --user --name=rapids_singelcell
 ```
 After you set up the enviroment you can install this package from this wheel into the enviroment. The wheel doesn't install any dependencies
 ```
-pip install https://github.com/Intron7/rapids_singlecell/releases/download/v0.2.1/rapids_singlecell-0.2.1-py3-none-any.whl
+pip install https://github.com/Intron7/rapids_singlecell/releases/download/v0.2.2/rapids_singlecell-0.2.2-py3-none-any.whl
 ```
 
 With this enviroment, you should be able to run the notebooks. So far I have tested these Notebooks on an A100 80GB, a Quadro RTX 6000 and a RTX 3090.
 
+To view a full guide how to set up a fully functioned single cell GPU accelerated conda environment visit [GPU_SingleCell_Setup](https://github.com/Intron7/GPU_SingleCell_Setup)
+
 ## Citation
 
 If you use this code, please cite: [![DOI](https://zenodo.org/badge/364573913.svg)](https://zenodo.org/badge/latestdoi/364573913)
diff --git a/rapids_singlecell/__init__.py b/rapids_singlecell/__init__.py
index 3983a6cf..172efc43 100755
--- a/rapids_singlecell/__init__.py
+++ b/rapids_singlecell/__init__.py
@@ -1,4 +1,4 @@
 from . import cunnData
 from . import scanpy_gpu_funcs
 
-__version__ = '0.2.1'
\ No newline at end of file
+__version__ = '0.2.2'
diff --git a/rapids_singlecell/cunnData.py b/rapids_singlecell/cunnData.py
index e059fe47..e60b7021 100755
--- a/rapids_singlecell/cunnData.py
+++ b/rapids_singlecell/cunnData.py
@@ -536,6 +536,7 @@ def normalize_pearson_residuals(self,
         X = X - mu
         X = X / cp.sqrt( mu + mu**2 / theta)
         X = cp.clip(X, a_min=-clip, a_max=clip)
+        X = cp.array(X, dtype= cp.float32)
         if inplace == True:
             if layer:
                 self.layers[layer]= X
@@ -1433,4 +1434,4 @@ def _check_nonnegative_integers(X):
     elif cp.any(~cp.equal(cp.mod(data, 1), 0)):
         return False
     else:
-        return True
\ No newline at end of file
+        return True
diff --git a/rapids_singlecell/scanpy_gpu_funcs.py b/rapids_singlecell/scanpy_gpu_funcs.py
index c384d9bb..7a2bd81b 100755
--- a/rapids_singlecell/scanpy_gpu_funcs.py
+++ b/rapids_singlecell/scanpy_gpu_funcs.py
@@ -22,6 +22,8 @@
 from cuml.cluster import KMeans
 from cuml.decomposition import PCA
 
+from scipy.sparse import issparse
+import warnings
 
 
 def _select_groups(labels, groups_order_subset='all'):
@@ -345,6 +347,11 @@ def pca(adata: AnnData,
          covariance matrix.
     """
     X = adata.layers[layer] if layer is not None else adata.X
+    if issparse(X):
+        warnings.warn(
+            "Your Countmatrix seems to be sparse, this can lead to a massive performance penalty.",
+            UserWarning,
+        )
     pca_func = PCA(n_components=n_comps, output_type="numpy")
     adata.obsm["X_pca"] = pca_func.fit_transform(X)
     adata.uns['pca'] ={'variance':pca_func.explained_variance_, 'variance_ratio':pca_func.explained_variance_ratio_}
@@ -907,4 +914,4 @@ def harmony_integrate(
     from . import harmonpy_gpu
     harmony_out = harmonpy_gpu.run_harmony(adata.obsm[basis], adata.obs, key, **kwargs)
 
-    adata.obsm[adjusted_basis] = harmony_out.Z_corr.T.get()
\ No newline at end of file
+    adata.obsm[adjusted_basis] = harmony_out.Z_corr.T.get()