Create utils.py

"I'm happy to support a pull request" ok

zodiax/experimental/utils.py

@@ -0,0 +1,34 @@
+import numpy as np 
+import scipy.cluster.hierarchy as sch
+import matplotlib.pyplot as plt 
+
+def cov_to_corr(cov):
+    # Calculate the standard deviations
+    std = np.sqrt(np.diag(cov))
+
+    # Calculate the correlation matrix
+    corr = cov / np.outer(std, std)
+
+    return corr
+
+def group_corr_inds(corr):
+
+    # print(corr)
+    # X = df.corr().values
+    X = corr 
+    # print(X.shape)
+    d = sch.distance.pdist(X)   # vector of ('55' choose 2) pairwise distances
+    L = sch.linkage(d, method='complete')
+    ind = sch.fcluster(L, 0.5*d.max(), 'distance')
+    return ind
+
+def visualize_covmat(cov,plot_names):
+    corr = cov_to_corr(cov)
+    ind = group_corr_inds(corr)
+
+    args = np.argsort(ind)
+    sorted_labels = [plot_names[i] for i in args]
+    plt.imshow(corr[args, :][:, args], origin="upper")
+    plt.xticks(range(len(plot_names)), sorted_labels, rotation=45)
+    plt.yticks(range(len(plot_names)), sorted_labels)
+    plt.colorbar()