diff --git a/HISTORY.rst b/HISTORY.rst
index d303acfd..e9423fcd 100644
--- a/HISTORY.rst
+++ b/HISTORY.rst
@@ -4,10 +4,10 @@ History
 
 X.Y.Z (YYYY-MM-DD)
 ------------------
+* Support missing data during BDA (:pr:`252`)
 * Document Fused RIME (:pr:`270`)
 * Add Multiton, LazyProxy and LazyProxyMultiton patterns (:pr:`269`)
 
-
 0.3.2 (2022-13-01)
 ------------------
 * Support numba >= 0.54 (:pr:`264`)
diff --git a/africanus/averaging/bda_mapping.py b/africanus/averaging/bda_mapping.py
index 34d5de62..a89d0561 100644
--- a/africanus/averaging/bda_mapping.py
+++ b/africanus/averaging/bda_mapping.py
@@ -303,8 +303,8 @@ def finalise_bin(self, auto_corr, uvw, time, interval,
             # The following is copied from DDFacet. Variables names could
             # be changed but wanted to keep the correspondence clear.
             # BH: I strongly suspect this is wrong: see eq. 18-19 in SI II
-            delta_nu = (lightspeed / (2*np.pi)) * \
-                       (self.decorrelation / max_abs_dist)
+            delta_nu = ((lightspeed / (2*np.pi)) *
+                        (self.decorrelation / max_abs_dist))
 
             fracsizeChanBlock = delta_nu / chan_width