BUG: Saving normalisation result in float (explicit upcasting)

Author: abcnishant007

lib/matplotlib/axes/_axes.py

@@ -2116,7 +2116,7 @@ def xcorr(self, x, y, normed=True, detrend=mlab.detrend_none,
         correls = np.correlate(x, y, mode="full")
 
         if normed:
-            correls /= np.sqrt(np.dot(x, x) * np.dot(y, y))
+            correls = correls / np.sqrt(np.dot(x, x) * np.dot(y, y))
 
         if maxlags is None:
             maxlags = Nx - 1