This project has moved. For the latest updates, please go here.

WeibullDistribution.FitToSample

Nov 29, 2012 at 3:31 PM

With some datasets, when I call WeibullDistribution.FitToSample(dat) it neither fails (in the Try/Catch/End Try construct) nor returns.

This is the sample data:

-        dat    {Meta.Numerics.Statistics.Sample}    Meta.Numerics.Statistics.Sample
        Count    5    Integer
+        InterquartileRange    {[0.00654597869813347,0.00679064783511837]}    Meta.Numerics.Interval
        IsReadOnly    False    Boolean
        Maximum    0.0088268887026073974    Double
        Mean    0.0069429346402411886    Double
        Median    0.0066506006265792872    Double
        Minimum    0.0059005573387674124    Double
        Name    "Curve Fit Data"    String
+        PopulationMean    {0.00694293464024119 ± 0.000495078671236537}    Meta.Numerics.UncertainValue
+        PopulationStandardDeviation    {0.00110702956309517 ± 0.000271065283501905}    Meta.Numerics.UncertainValue
+        PopulationVariance    {1.22551445356668E-06 ± 6.00154564730763E-07}    Meta.Numerics.UncertainValue
        StandardDeviation    0.00099015734247307461    Double
        Variance    0.00000098041156285334141    Double
+        Results    Expanding will process the collection    <enumerable type>

Coordinator
Nov 29, 2012 at 7:15 PM

Thanks for the error report! I have been able to reproduce and have created a bug (http://metanumerics.codeplex.com/workitem/7213). Will get back to you as soon as I have had a chance to investigate.

Nov 29, 2012 at 7:17 PM
You are welcome. Look forward to the fix and hopefully the Gamma fit routine.

Thomas E. Wierman
Risk Assessment and Management Services
208 526-3836 (office)
208 221-4865 (mobile)



On Thu, Nov 29, 2012 at 1:15 PM, ichbin <notifications@codeplex.com> wrote:

From: ichbin

Thanks for the error report! I have been able to reproduce and have created a bug (http://metanumerics.codeplex.com/workitem/7213). Will get back to you as soon as I have had a chance to investigate.

Read the full discussion online.

To add a post to this discussion, reply to this email (metanumerics@discussions.codeplex.com)

To start a new discussion for this project, email metanumerics@discussions.codeplex.com

You are receiving this email because you subscribed to this discussion on CodePlex. You can unsubscribe on CodePlex.com.

Please note: Images and attachments will be removed from emails. Any posts to this discussion will also be available online at CodePlex.com


Coordinator
Dec 6, 2012 at 8:58 AM

The 2.1.0 distribution contains this bug but it was fixed by a subsequent check-in (http://metanumerics.codeplex.com/SourceControl/changeset/70401) that contained an improved fit algorithm. Not only does the new algorithm not have the bug, but it is also faster. The same check-in also contains a GammaDistribution.FitToSample method.

Until the next official release, you will have to do your own build to get these changes. If you sync only GammaDistribution.cs and WeibullDistribution.cs, you will get the two FitToSample methods you want with the minimum number of other changes. At the moment, we are not expecting to do a 2.1.1 bug-fix release (and of course if we did, it would only have the Weibull fix, not the new Gamma fit method) before the 2.2.0 release, although that might change if enough serious bugs accumulate.

Dec 6, 2012 at 6:13 PM
Thanks for the new code. However, I have not got a clue on how to "sync" the check-ins. Are there any instructions on how to do this?

Thomas E. Wierman
Risk Assessment and Management Services
208 526-3836 (office)
208 221-4865 (mobile)



On Thu, Dec 6, 2012 at 2:58 AM, ichbin <notifications@codeplex.com> wrote:

From: ichbin

The 2.1.0 distribution contains this bug but it was fixed by a subsequent check-in (http://metanumerics.codeplex.com/SourceControl/changeset/70401) that contained an improved fit algorithm. Not only does the new algorithm not have the bug, but it is also faster. The same check-in also contains a GammaDistribution.FitToSample method.

Until the next official release, you will have to do your own build to get these changes. If you sync only GammaDistribution.cs and WeibullDistribution.cs, you will get the two FitToSample methods you want with the minimum number of other changes. At the moment, we are not expecting to do a 2.1.1 bug-fix release (and of course if we did, it would only have the Weibull fix, not the new Gamma fit method) before the 2.2.0 release, although that might change if enough serious bugs accumulate.

Read the full discussion online.

To add a post to this discussion, reply to this email (metanumerics@discussions.codeplex.com)

To start a new discussion for this project, email metanumerics@discussions.codeplex.com

You are receiving this email because you subscribed to this discussion on CodePlex. You can unsubscribe on CodePlex.com.

Please note: Images and attachments will be removed from emails. Any posts to this discussion will also be available online at CodePlex.com


Dec 18, 2012 at 5:36 PM
Looking at your web site, it appears that there are several versions of the project. How can you tell which is the one that the changes should be applied to? Is it the Numerics v1 Core?

Thomas E. Wierman
Risk Assessment and Management Services
208 526-3836 (office)
208 221-4865 (mobile)



On Thu, Dec 6, 2012 at 12:13 PM, Wierman, Thomas E <thomas.wierman@inl.gov> wrote:
Thanks for the new code. However, I have not got a clue on how to "sync" the check-ins. Are there any instructions on how to do this?

Thomas E. Wierman
Risk Assessment and Management Services
208 526-3836 (office)
208 221-4865 (mobile)



On Thu, Dec 6, 2012 at 2:58 AM, ichbin <notifications@codeplex.com> wrote:

From: ichbin

The 2.1.0 distribution contains this bug but it was fixed by a subsequent check-in (http://metanumerics.codeplex.com/SourceControl/changeset/70401) that contained an improved fit algorithm. Not only does the new algorithm not have the bug, but it is also faster. The same check-in also contains a GammaDistribution.FitToSample method.

Until the next official release, you will have to do your own build to get these changes. If you sync only GammaDistribution.cs and WeibullDistribution.cs, you will get the two FitToSample methods you want with the minimum number of other changes. At the moment, we are not expecting to do a 2.1.1 bug-fix release (and of course if we did, it would only have the Weibull fix, not the new Gamma fit method) before the 2.2.0 release, although that might change if enough serious bugs accumulate.

Read the full discussion online.

To add a post to this discussion, reply to this email (metanumerics@discussions.codeplex.com)

To start a new discussion for this project, email metanumerics@discussions.codeplex.com

You are receiving this email because you subscribed to this discussion on CodePlex. You can unsubscribe on CodePlex.com.

Please note: Images and attachments will be removed from emails. Any posts to this discussion will also be available online at CodePlex.com



Coordinator
Feb 11, 2013 at 10:45 AM
The path to the specific file you want is Numerics/Core/Core/Statistics/Distributions/WeibullDistribution.cs. Numerics-v1 is an old release branch containing the last snapshot of the version 1.x code.
Feb 11, 2013 at 2:26 PM
<div dir="ltr"> <div> <div>You should know that I have never been part of an open source project. Therefore, I do not have the solution running on my computer.<br> <br> </div> I downloaded metanumerics-71216 and copied the Numerics directory to my machine and tried to open the Numerics solution. It wants to connect to <a href="http://tfs10.codeplex.com">tfs10.codeplex.com</a> with a user name and password.<br> <br> </div> All stop here. What to do?<br> </div> <div><br clear="all"> <div> <div><span style="font-family:'arial black',sans-serif"><span></span><span></span><img src="http://i.imgur.com/KTgVz.gif"> Thomas E. Wierman</span></div> <div><b><span style="color:rgb(51,102,255)">Risk Assessment and Management Services</span></b> </div> <div><i>208 526-3836 (office)</i></div> <div><i>208 221-4865 (mobile)</i></div> </div> <br> <br> <div>On Mon, Feb 11, 2013 at 4:45 AM, ichbin <span dir="ltr">&lt;<a href="mailto:notifications@codeplex.com" target="_blank">notifications@codeplex.com</a>&gt;</span> wrote:<br> <blockquote style="margin:0 0 0 .8ex; border-left:1px #ccc solid; padding-left:1ex"> <div> <p>From: ichbin</p> <div>The path to the specific file you want is Numerics/Core/Core/Statistics/Distributions/WeibullDistribution.cs. Numerics-v1 is an old release branch containing the last snapshot of the version 1.x code.<br> </div> <div> <p>Read the <a href="http://metanumerics.codeplex.com/discussions/405042#post999521" target="_blank"> full discussion online</a>.</p> <p>To add a post to this discussion, reply to this email (<a href="mailto:metanumerics@discussions.codeplex.com?subject=[metanumerics:405042]" target="_blank">metanumerics@discussions.codeplex.com</a>)</p> <p>To start a new discussion for this project, email <a href="mailto:metanumerics@discussions.codeplex.com" target="_blank"> metanumerics@discussions.codeplex.com</a></p> <p>You are receiving this email because you subscribed to this discussion on CodePlex. You can <a href="https://metanumerics.codeplex.com/discussions/405042/unsubscribe/" target="_blank"> unsubscribe</a> on CodePlex.com.</p> <p>Please note: Images and attachments will be removed from emails. Any posts to this discussion will also be available online at CodePlex.com</p> </div> </div> </blockquote> </div> <br> </div>
Feb 12, 2013 at 2:37 PM
<div dir="ltr"> <div>I think I figured it out. I just compiled the 71216. Hope that's correct.<br> <br> Now Weibull, Gamma, Lognormal, Beta, Exponential, and Normal all seem to work.<br> <br> </div> Thanks.<br> </div> <div><br clear="all"> <div> <div><span style="font-family:'arial black',sans-serif"><span></span><span></span><img src="http://i.imgur.com/KTgVz.gif"> Thomas E. Wierman</span></div> <div><b><span style="color:rgb(51,102,255)">Risk Assessment and Management Services</span></b> </div> <div><i>208 526-3836 (office)</i></div> <div><i>208 221-4865 (mobile)</i></div> </div> <br> <br> <div>On Mon, Feb 11, 2013 at 8:26 AM, Wierman, Thomas E <span dir="ltr">&lt;<a href="mailto:thomas.wierman@inl.gov" target="_blank">thomas.wierman@inl.gov</a>&gt;</span> wrote:<br> <blockquote style="margin:0 0 0 .8ex; border-left:1px #ccc solid; padding-left:1ex"> <div dir="ltr"> <div> <div>You should know that I have never been part of an open source project. Therefore, I do not have the solution running on my computer.<br> <br> </div> I downloaded metanumerics-71216 and copied the Numerics directory to my machine and tried to open the Numerics solution. It wants to connect to <a href="http://tfs10.codeplex.com" target="_blank">tfs10.codeplex.com</a> with a user name and password.<br> <br> </div> All stop here. What to do?<br> </div> <div><br clear="all"> <div> <div><span style="font-family:'arial black',sans-serif"><span></span><span></span><img src="http://i.imgur.com/KTgVz.gif"> Thomas E. Wierman</span></div> <div><b><span style="color:rgb(51,102,255)">Risk Assessment and Management Services</span></b> </div> <div><i>208 526-3836 (office)</i></div> <div><i>208 221-4865 (mobile)</i></div> </div> <div> <div><br> <br> <div>On Mon, Feb 11, 2013 at 4:45 AM, ichbin <span dir="ltr">&lt;<a href="mailto:notifications@codeplex.com" target="_blank">notifications@codeplex.com</a>&gt;</span> wrote:<br> <blockquote style="margin:0 0 0 .8ex; border-left:1px #ccc solid; padding-left:1ex"> <div> <p>From: ichbin</p> <div>The path to the specific file you want is Numerics/Core/Core/Statistics/Distributions/WeibullDistribution.cs. Numerics-v1 is an old release branch containing the last snapshot of the version 1.x code.<br> </div> <div> <p>Read the <a href="http://metanumerics.codeplex.com/discussions/405042#post999521" target="_blank"> full discussion online</a>.</p> <p>To add a post to this discussion, reply to this email (<a href="mailto:metanumerics@discussions.codeplex.com?subject=[metanumerics:405042]" target="_blank">metanumerics@discussions.codeplex.com</a>)</p> <p>To start a new discussion for this project, email <a href="mailto:metanumerics@discussions.codeplex.com" target="_blank"> metanumerics@discussions.codeplex.com</a></p> <p>You are receiving this email because you subscribed to this discussion on CodePlex. You can <a href="https://metanumerics.codeplex.com/discussions/405042/unsubscribe/" target="_blank"> unsubscribe</a> on CodePlex.com.</p> <p>Please note: Images and attachments will be removed from emails. Any posts to this discussion will also be available online at CodePlex.com</p> </div> </div> </blockquote> </div> <br> </div> </div> </div> </blockquote> </div> <br> </div>