« first day (2289 days earlier)   

12:54 AM
@CarlLange - I don't always understand Dataset syntax, but I think you want to put Accumulate in the first argument, instead of All
`ds = Dataset@
MapThread[
Association["val1" -> #1, "val2" -> #2] &, {{1, 8, 8, 5,
6}, {Quantity[3, "Feet"], Quantity[0, "Feet"],
Quantity[0, "Feet"], Quantity[5, "Feet"], Quantity[3, "Feet"]}}]`
ds[Accumulate, "val2"]
(I forgot to turn my notebook magnification down to 50% before screenshots)
 
1:14 AM
@CarlLange ds[All, i = 0; {"RateChange" -> (i += # &)}] keeps the dates but it does leave i hanging around; which I don't like
 
 
3 hours later…
4:37 AM
Sigh. TimeSeriesResample was very broken for me, until Mma decided to do a paclet update. I wonder how often this kind if things happen.
 
 
5 hours later…
9:18 AM
@kirma there should be a function to call a dataset with the current version installed, all Wolfram packet updates added on top of it, a all others that might exist, but still not installed...
 
I already started to write a bug report on it, but then restarted the kernel, and next evaluation did something that looked like a paclet update... and the problem was gone.
 
 
1 hour later…
 
6 hours later…
4:50 PM
I need some advice from Mathematica users who speak English natively.
Why is natural as a function name? TreeLikeComponents or TreelikeComponents?
For those who wonder what these "tree-like components" are, see below:
@JasonB. You can use Edit -> Copy As -> Bitmap on a retina machine to get a non-retina rasterization into the clipboard.
 
5:12 PM
@Szabolcs I slightly prefer TreelikeComponents
 
@chuy Agreed, camel-case indicates to me that you have tree like components instead of treelike components. The second better conveys what you want, even if treelike isn't truly tree-like
It's English, so smushing words together like that is perfectly valid outside of very formal settings.
 
@kirma TimeSeriesResample leaks memory. I reported it but I haven't checked if it was solved in 11.3
 
@gdelfino That's better than basically failing to handle any time quantities (or implied time quantities it generated itself) as the second argument. That was what I saw.
 
5:28 PM
@chuy @b3m2a1 Thanks! That's what I had named it, just wanted confirmation before release.
 
 
2 hours later…
7:52 PM
@JasonB. Ah, that's really helpful, thanks a lot! Unfortunately for me it's lost the Quantity, but that's still far nicer than what I had before!
@Edmund I suppose if you do it in a Module it won't leave i hanging around: ds[All, Module[{i = 0}, {"RateChange" -> (i += # &)}]] works correctly (although loses the Quantity for me. Thanks a lot though, that's really helpful!
 
8:18 PM
@JasonB. that's funny, doing ecbRateChanges[Accumulate, {"RateChange"}] correctly keeps the Quantity. Weird.
 
 
2 hours later…
9:55 PM
@Szabolcs About your VertexComponent question. Have you tried what happens when you use DirectedEdge? This result looks suspicious as well:
g3 = Graph[edges /. Rule :> DirectedEdge, VertexLabels -> Automatic];
HighlightGraph[g3, Subgraph[g, VertexComponent[g3, {50}, 3]],
   GraphHighlightStyle -> "Thick"]
How do I come from 1 to 50 in 3 steps?
 

« first day (2289 days earlier)