(bug) numpy 2.0 is not supporting np.NaN

weikang9009 · weikang9009 · commit dca74cc281a8 · 2024-11-02T11:31:00.000-07:00
diff --git a/segregation/singlegroup/absolute_centralization.py b/segregation/singlegroup/absolute_centralization.py
@@ -143,8 +143,8 @@ def _absolute_centralization(
     Xi = np.cumsum(x[asc_ind]) / X
     Ai = np.cumsum(area[asc_ind]) / A
 
-    ACE = np.nansum(shift(Xi, 1, cval=np.NaN) * Ai) - np.nansum(
-        Xi * shift(Ai, 1, cval=np.NaN)
+    ACE = np.nansum(shift(Xi, 1, cval=np.nan) * Ai) - np.nansum(
+        Xi * shift(Ai, 1, cval=np.nan)
     )
 
     core_data = data[[group_pop_var, total_pop_var, data.geometry.name]]

Original file line number	Diff line number	Diff line change
`@@ -143,8 +143,8 @@ def _absolute_centralization(`
`143`	`143`	`Xi = np.cumsum(x[asc_ind]) / X`
`144`	`144`	`Ai = np.cumsum(area[asc_ind]) / A`
`145`	`145`
`146`		`- ACE = np.nansum(shift(Xi, 1, cval=np.NaN) * Ai) - np.nansum(`
`147`		`- Xi * shift(Ai, 1, cval=np.NaN)`
	`146`	`+ ACE = np.nansum(shift(Xi, 1, cval=np.nan) * Ai) - np.nansum(`
	`147`	`+ Xi * shift(Ai, 1, cval=np.nan)`
`148`	`148`	`)`
`149`	`149`
`150`	`150`	`core_data = data[[group_pop_var, total_pop_var, data.geometry.name]]`