I gave a talk at an embedded workshop (Hybrid Symbolic-Numeric Computation) while at ICIAM. But the talk Sylvia noted was at Beijing Jiaotong Technical University. On a different note,
(I seem to have issues with chat...where was I...) I gave a talk at an embedded workshop (Hybrid Symbolic-Numeric Computation) while at ICIAM. But the talk Sylvia noted was at Beijing Jiaotong Technical University. On a different note, I saw mention of tossing a baby from a roof. I don't recall that being a common notion and was wondering: did someone else here read early issues of "Cerebus the Aardvark"?
Also saw mention of shucking and eating live shellfish in Japan. This is probably still done in the States, with oysters at least. Was definitely around in Boston's Hay Market, as recently as the 80's if not still today. Lewis Carroll's Walrus would have approved...
Ideally, the background would cover the exact same width as the one above and below, which are in style "Text", with the extra indentation due to itemness inside the backgroun rectangle
But I am turning greedy I guess. I've been wondering why mma paints backgrounds on lists like that for over a decade :p
You won't have to play with them very much. If you drill into Defaults.nb, you can find the actual numbers that are used on your system for StyleData["Text"]. I'm guessing they're the same across platforms/versions but in case they're not, just use those for Item as well
@MarianoSuárez-Alvarez You're not copying things around. Insertion/deletion is cheap on associations but I don't think they're in place. "In place" would imply mutation whereas associations are immutable
I. General
I will first try to briefly answer the questions, and then illustrate this with a small but practical application.
1.Speed of insertion / deletion
Associations are based on so called Hash Array Mapped Trie persistent data structure. One can think of this as a nested hash table, but ...
Streaming` module - general, and the case at hand
Starting with V10.1, there is an undocumented support for certain lazy operations in Mathematica. However, the primary goal of Streaming` is to support out of core computations reasonably efficiently, and lazy operations are only the secondary go...
Experimental/undocumented, etc. but something to play with
Does anyone know if the following difference is by design and should be expected? StringCases["kk", "k" ~~ ___] --> {"kk"} StringCases["kk", "k*"] --> {}
@IstvánZachar I don't know if the inconsistency is helpful, but it is documented as such. The docs for StringMatchQ specifically call out the special characters, but StringCases doesn't.
@MichaelHale I'm not sure I see how the StringMatchQ doc reveals anything about this, but I might have simply overlooked it. Could you please point me to the specific line in the docs? Actually, StringMatchQ["kk","k"~~___] equals StringMatchQ["kk","k*"], so I assume that the discrepancy is in the way StringCases partitions the string.
@MichaelE2 Well, can you imagine having a c-source with 10^6 lines of code only for the initialization? I guess if I were a compiler I would be pissed too :-)
@MichaelE2 When you do something like this Compile[{{m, _Integer, 2}... then you can never change m because it is not copied. Mathematica uses the pointer to the real m and therefore, you can just pass the very large matrix as argument and you are fine.
@halirutan Yes, that was my solution. My thought was to compare the two for the OP. Only compiling the included version slowed me way down. Talk about pissed software. The FE would become unresponsive for a minute or two every now and then, too.
Yet somehow I think a compiler should handle 10^6 20-byte or so strings for reals without too much fuss.
@MichaelE2 I did a mistake by trying to compute all g3 for all Tuples[Range[1000],{2}]... whole system was gone until I ssh`d in and killed the kernel :-)
