DSkoog

Daniel Skoog

1077 Reputation

15 Badges

6 years, 83 days
Maplesoft
Maple Product Manager

Social Networks and Content at Maplesoft.com

Maple Application Center
Daniel Skoog is a Product Manager for Maplesoft. He holds a B.Sc. in Mathematics from Queen’s University, Canada and a M.Sc. in Financial Mathematics from Uppsala University, Sweden. Daniel joined Maplesoft in 2011 and has used his applied technical expertise of mathematics, statistics, data analysis and software design to deliver several exciting new releases. Daniel has also contributed several commands to the Maple library, having authored new routines in Statistics, Finance and data analysis.

MaplePrimes Activity


These are replies submitted by DSkoog

Building on @Carl Love's answer, you may also be interested in quickly viewing a summary for the model using the summarize option (Maple 2016+):

X:= <21.2, 24.7, 20.8, 20.8, 20.3>:
Y:= <16.6, 19.7, 16.4, 16.8, 16.9>:
L:= Statistics:-LinearFit([1,x], <X|Y>, x, summarize);

Returns:

Summary:
----------------
Model: 1.4368206+.73484134*x
----------------
Coefficients:
              Estimate  Std. Error  t-value  P(>|t|)
Parameter 1    1.4368    2.6933      0.5335   0.6307
Parameter 2    0.7348    0.1246      5.8985   0.0097
----------------
R-squared: 0.9206, Adjusted R-squared: 0.8942

@Christopher2222 I've filed an internal bug for this issue. Thanks for bringing it to our attention.

@tomleslie : I can't speak for the rest, but I'm not sure if I'd say none of the systems want to admit that they are using LAPACK, BLAS, etc. Check out the section 'The Computation Engine' in the first chapter of the Maple programming guide: https://www.maplesoft.com/support/help/Maple/view.aspx?path=ProgrammingGuide/Chapter01 


I would definitely be in favour of this. This has been mentioned a few times ( see http://www.mapleprimes.com/posts/207108-New-Look-Of-MaplePrimes#comment201900 for some more recent comments), though I don't think anyone has actually gone through the process of creating the proposal on Area 51, so it's great that you did.

Personally I like both MaplePrimes and Stack Exchange for different reasons. I find myself on both sites daily. In case it's useful for anyone else, I created a twitter feed that aggregates posts from both MaplePrimes, all the Stack exchange pages and the Google group for Maple here:

https://twitter.com/MPLProgramming

Just in case anyone didn't see it, NVIDIA has posted a fix: https://nvidia.custhelp.com/app/answers/detail/a_id/4378  So updating your NVIDIA driver to 378.57 or above should resolve the issues.

@Henjue : We've been looking into this issue and found the following thread on Reddit: https://www.reddit.com/r/feedthebeast/comments/5pydv9/dont_install_nvidia_gpu_driver_version_37849_it/ It would appear that NVIDIA is working on a hotfix. We will still continue looking into it from our end too, but it may be beneficial to hold off updating to the latest NVIDIA driver until a newer hotfix is available.

If it helps, here's a png version of the help icon that ships with the Windows version. It is 256x256, so you might need to rescale it.

@Christopher2222 : Unless I'm missing something, if you are looking to collect data with "attributes" (meta-information), then a DataFrame might be another alternative. DataFrames store meta-data as column and row names and can also handle specifying a datatype for each column of information. Moreover, they print in a table layout (similar to matrices). For example:

CarInfo := DataFrame(<<Honda, Honda, Honda>|<green,red,blue>|<undefined,Civic,CRV>|<2008,2004,float(undefined)>|<undefined,undefined,`2WD`>>, columns = <Make, Color, Model, Year, `2/4WD`>);

CarInfo[1,Model] := Accord; #Add in the missing information
CarInfo;

The resulting DataFrame would look like:

@Christopher2222 : Sorry, I spoke too soon. The two-argument calling sequence for GetDocumentProperty actually isn't available yet. It will be available in Maple 2017.

For now, you can only use the GetDocumentProperty command to interogate the current document.

@nm : I'd be lying if I said that this hadn't crossed my mind several times. There are already many MaplePrimes users that do frequent the Stack Exchange pages (mostly the Stack Overflow and Mathematics sites). I even have a filter that checks for new "Maple" related posts across all stack exchange sites: http://stackexchange.com/filters/247545/maple

I don't see any major reason why as a group of Maple users, we couldn't propose a new Maple Stack Exchange site, however the problem is simply that to get a new Stack Exchange site up and running takes a lot of motivated users. See here for more details on the process of staging a new site: http://area51.stackexchange.com/faq

Personally, I'm happy to see both exist, as I do feel like they offer slightly different features and experiences. In fact, I do often see a lot of other stack users redirecting questions to MaplePrimes when an 'expert' answer is required, which speaks volumes about our community here.

@acer : Yes, you're correct, I had misread the original post. So we'll file this as shortcoming in the numeric formatting mechanism.

@David Moss Thanks for submitting this post. I suggested to David to start this post to generate exactly this kind of dialogue.

@Carl Love I can definitely see where you're coming from and do ultimately agree that we should tend toward to having fewer and more intuitive commands whenever possible.

The case that you mention does definitely work, but I was also concerned with other cases like:

DS:=DataSeries(<"a","b">,datatype=string);
DataSeries(DS,datatype=name);

returns an error, whereas the following works:

DataSeries(((x)->convert(x,name))~(DS),datatype=name);

Now I'm guessing that you would argue that we should just make the DataSeries( foo, datatype=type) just work, which would be nice and very simple, but I'm still not convinced that it should be the job of the DataSeries command itself to change the data into a given datatype.

Using the Matrix example:

M:=Matrix(<<1,2>|<"a","b">>,datatype=anything);
Matrix(M,datatype=string);

Returns an error. This is similar to the DataSeries constructor in that it does not attempt to change the data into the different datatype.

This will definitely be a good point to debate once beta season rolls around...

@Carl Love : While it is pretty straight forward to accomplish this for string -> numeric, I think that it is rather desirable for users to have a more general command that can change entries in a DataSeries from one datatype to an another arbitrary datatype. For example, parse will work in the case of going from string to numeric, but we would need to map convert or some other command onto the entries in order to, say, change the datatype of the entries from string to name.

Now, this functionality is actually something that I've already added into Maple 2017. We wanted to make it as easy as possible for users to instead just specify Datatype( DataSeries, type ) and have Maple work out the rest. That said, the primary function of this new function is actually going to be a little different than what I've said above: it first changes the indicated datatype of the DataSeries, and then attempts to coerce the entries in the DataSeries to that datatype.

This is probably a better indication of what the planned command will do:

DS := DataSeries( [1,2,3], datatype = integer );
SubsDatatype( DS, float );

Output:

DataSeries( [1.,2.,3.], datatype = float );

@Mariusz Iwaniuk : You might also find this plot interesting:

Explore(plot(([Re, Im])(x^k), x = -2 .. 2, view = [-2 .. 2, -3 .. 3], legend = ["Re", "Im"]), k = 1.0 .. 3)
1 2 3 4 5 6 7 Page 1 of 7