I stopped using Python in my research, too. Now I'm using R which is so terrible as a language that for the first two years I felt insulted every time I wrote

I stopped using Python in my research, too. Now I'm using R which is so terrible as a language that for the first two years I felt insulted every time I wrote code in it. However, I stuck with it and even like it now because it has an amazing ecosystem. (I also got used to expressing things in terms of vector and matrix operations.) Julia is promising but things are still too much in flux for my taste to consider it for serious work. I have no doubts about Julia as a language but I'm not sure if they are going to able to build an ecosystem like R's. They're not always making it easy for R developers to switch. Fortunately, some R veterans like Doug Bates are heavily investing in Julia. I hope others will follow.