![]() which is good to have access to all that computer power in the uni, but the waiting suks. pretty much all the sciences gather there with their scripts to be ran and get their results. all the engineers, physics students, chem, math/stats guys. i was trying to avoid that because, unless one gets early, they get kinda crowded. i havent even been able to try the thing suggested before to see it works because i dont wanna have R stop wherever it is.ĭarn, i guess i'm gonna have to go to one of the computer labs get one of those massively powerful computers to run this script for me. it's 6:30pm here in rainy vancouver and the thing is STILL running. and i did try with an easier example, but i guess i made it too easy and didnt get any insight as for how long this was going to take. That leaves a lot of room for creativity, but when you want to solidify those procedures, then an S4 OOP approach can provide that.Ĭlick to expand.with regards to your first question dason, no, i didnt and that was very, very stoopid of me because it would've only taken one line of code more. R is at its heart a functional programming language, which is great. In the end, I think it is a question of structure vs functionality. Now, if you are an S4 programmer developing certain R routines, it makes all the sense in the world to do it that way, or if you want to maintain consistent behavior when it ports between other technologies (e.g., I would look to the S4 implementation if I were using it as the foundation for an ArcGIS statistical analysis that could be used broadly so that it coordinates with Python (the port) in a specific manner). But that is not really the "R way," which is why it begs a serious question. In that case, you **** well want your S4 objects to behave in an intended way and to be accessed only in a specified way. I think it would really take off if R developed specific and rigorous "this is how you do this. One could technically create their own, but getting involved at all begs to question why one would use the S4 approach anyway. Yeah Dason, I don't expect there to be too much in the away of getter methods, but R just isn't an object-oriented language to begin with. ![]() any ideas on how to get (free) extra processing power? please, let me know. ![]() i'd say i can hold my ground in the field of computer-science-related stuff, but i'm really not as skilled as most people are. a friend of mine told me i should try and alternate between EM and Newton-Raphson, another one told me that lets you use their servers for free in cases like this one. does anyone know how to do this a little bit faster? i was considering running the simulation at least 1000 times and i dont wanna die of old age until it finishes. ![]() i thought it would be faster but there's like a sub for-loop where it runs 5 times and then a "master" for-loop (so to speak) for 100 times, so i guess you could say it's being run 500 times (which i think it's a very small number for a simulation). it's basically doing the EM algorithm over and over again. it actually started today around 10am pacific time and right now at 2:30pm it's still running. my simulation is taking AAAGES to finish. i'm very, VERY prone to spaghetti codingĪnyways, now that we're on topic i was wondering about something else (i'm not sure whether i should start another thread or not). my original simulation is a little more complicated so i only wanted to add the code that is giving me trouble. i think i had that "dat1" variable defined somewhere else.
0 Comments
Leave a Reply. |