11,478 reputation
12571
bio website linkedin.com/pub/andy-ross/14/…
location Columbia, MO
age 32
visits member for 2 years, 6 months
seen 10 hours ago

10h
comment FindDistributionParameters of a sum of a mixture distribution
To my knowledge you won't have any luck with ProbabilityDistribution if you don't have a closed form for one of the distribution functions.
Jul
18
comment Why do EstimatedDistribution and DistributionFitTest work this way
You should find the answer to this in my response to this question.
Jun
18
comment Why am I getting wildly incorrect results from FirstPassageTimeDistribution with inexact transition matrix?
This one wasn't my baby so I can't really say for sure what is going on. I suspect that a completely different algorithm is being used for inexact numbers and it is failing. You should file a bug.
May
19
comment Replicability of RandomVariate within a function?
Any reason you are avoiding Mean[Abs@Differences[x]]? That seems a lot simpler here. Not to mention, it will be much faster.
May
15
comment Undo differences in a time series
Right. ARIMA does all of this under the hood.
May
14
comment What is the meaning of Automatic in goodness of fit test?
@Szabolcs I don't see anything wrong with your approach. I don't have access to the source so I can't promise that the Infinity is safe across the board. You know the typical caveat about undocumented internals being subject to change.
May
14
comment Undo differences in a time series
I recommend fitting an ARIMA process to the original data. It includes the trend component so you won't have to undo the differencing.
May
2
comment SmoothHistogram with sharp boundary
I believe my answer to this question is what you are looking for.
Apr
20
comment Determining the dimension of a probability distribution
@Szabolcs I believe DistributionDomain is fairly robust but it may be that the representation of domains may change in the future. Some things are pretty wonky, like held ranges to infinity, and probably need different symbols to represent them.
Apr
17
comment Determining the dimension of a probability distribution
There is also Statistics`Library`DistributionDimensionality but I make no promises that it is robust.
Apr
2
comment Plot Wiener Process and its running maximum
Yes, that is correct.
Mar
31
comment Requesting explanation of FitModel code
I've gone to your chat room. that might be more efficient.
Mar
31
comment Requesting explanation of FitModel code
You should be using "Properties" not "Field". Also note that you need to use quotes.
Mar
31
comment Requesting explanation of FitModel code
Notice that the dispatch created in With contains a list of rules. The left-hand side of the rules are property names (in this case BreuschPagan and WhitesHeteroskedasticity) the right hand side are the values (here injected using With).
Mar
31
comment Requesting explanation of FitModel code
Indeed. I use these so frequently that I often take them for granted. I'll add a comment to this effect.
Mar
26
comment Inconsistent DistributionFitTest results?
@Steve see my edit
Mar
26
comment Inconsistent DistributionFitTest results?
To test goodness of fit I recommend looking at a QuantilePlot against the fitted distribution you give. It appears that it fits pretty well until we get out into the right tail.
Mar
26
comment Inconsistent DistributionFitTest results?
@Steve, sorry I was unsuccessfully trying to multitask. I've fixed the copy-paste errors.
Mar
25
comment What is the best distribution for my histogram?
Just for clarity, DistributionFitTest defaults to a test for normality. The fitted distribution is NOT the best fitting distribution. It is the best fitting distribution in the family you are testing. When setting the distribution family to Automatic you are using the normal family of distributions.
Mar
5
comment RandomSample Bug?
@Szabolcs I suspected auto-compilation as well.