Replies: 3 comments 2 replies
-
This is likely a typo in our documentation, which I have now corrected. The previous input deck would probably run, but the behaviour would be unpredictable. Good spot! The EPOCH maths parser only supports the |
Beta Was this translation helpful? Give feedback.
-
Thanks. Extending math parser could be even more the step into wrong direction. My point is that current IF construct complicates the My experience says that probably 20-30% of CPU time with my different codes is lost due to errors in setup. By the away Is IF FALSE part optional ? EPOCH probably has setup difficulty 8 out of 10. Not in the sense it is so dark and complex but that it is easy to make a move you did not intended and did not do what you intended to do. At least initially when you learn how to use it. All is kind of acceptable, though it is pity to lose a day and find the error next day. But in some cases it becomes a disaster when you got just one day in a year on 1 million cores machine and found next day that wasted it due to setup error. I am just learning and i already lost on this IF-True_False couple weeks. I used two IFs and the next one's False part interfered with the previous one ;( . It is the same like to give a small child complex task: go there and if you find this - then do this and if that then this. 50% chance is he will fail. We all are always like that small children every time with the new confusing task As to Epoch unjustified complexities, Epoch has initial data and output in one directory, EXE is in other and it starting from the third, all with specific names. If i want all to be on one arbitrary directory - what i need to do, ideally with just one click ? |
Beta Was this translation helpful? Give feedback.
-
I wrote before about IF construct of math parser (see #412 and after 2 years of using parser even more convinced that it was huge design defect to offer for it only one option if(A) Then B Otherwise C This "Otherwise C" should be optional like all languages allow. For example in Fortran you can write either
So my suggestion for developers would be to modify parser so that "Otherwise C" would be optional Because this "Otherwise C" makes setup a hell when multiple species are present. You scratch the head for hours and do multiple tries losing days because in supercomputers often see results only next day. You always have to figure out what to put into this "otherwise C" when you have 3-5 different species. And the solution looks totally unnatural and redundant. And even feels like being dumb and wrong Here is an example: suppose you need to setup 4-layers "cake" of different element That's all you need ! But with current method though For specie1 EPOCH is great and fast. Really worst part of EPOCH is this dumb IF |
Beta Was this translation helpful? Give feedback.
-
There density of electrons defined via density of Carbon which is defined after that. But density of Carbon is defined via density of electron. Am i losing something?
number_density = if(abs(x) lt thick, den_max)
Beta Was this translation helpful? Give feedback.
All reactions