New issue 1141: Problem to sample a ray in yt 3.2 https://bitbucket.org/yt_analysis/yt/issues/1141/problem-to-sample-a-ray-in-... Kuo-Chuan Pan: Dear all, I am not sure if here is the correct place to ask question (if not, please forgive me). Recently, I am updating my analysis codes from yt 2.x to 3.2. I found that the pf.h.ray method is very different in between yt 2.x and 3.x. I use pf.h.ray(point1, point2) in yt 2.x and ds.ray(point1,point2) in yt3.2. In yt 3.2, sometimes I got very strange results in ray('t'). The returned value is simply 0 1, instead of an array of sampled data. In addition, sometimes the returned array in ray('t') is not sorted in radius. I know that I could sort the array by using argsort() but is this due to performance issue or did I made some stupid mistake to use ds.ray()? Also, is this a bug when ds.ray(point1,point2) returns just 0 1? I am sure I am giving correct point1 and point2 and with a reasonable length. Many thanks, Kuo-Chuan