I'm not a Bayesian expert, so would welcome constructive feedback. At a minimum, it'll help me get ahead of questions at the seminar itself.

For a while now, I've been putting some data sets into a package called "raw" (r actuarial workshops) for use in the workshops that I've been a part of. It's always been hosted on GitHub, but I'd like to try to move it to CRAN. That's one less step for attendees to be able to load the package. At present, I've got the following data sets:

• All of the COTOR challenge claims data
• New Jersey Manufacturers WC triangle (from the recent Tayler and Grainne monograph)
• Tracks of every hurricane covered by the NOAA since 1972
• NFL results for about 12 seasons (I use this as an example of logistic regression)
• Some simulated data for state and regional claims experience (I use this for some ggplot2 examples and possibly for some hierarchical modeling examples)
• Most of the PPA data from Appendix A of the Werner & Modlin basic ratemaking paper

Is there anything that anyone would like to see added? Anything deleted? I think the NFL stuff is fun, but was thinking about switching it to a claims closure example.

I'm aiming for a CRAN submission sometime in late July or early August. Yesterday, I got package up on CRAN without any notes or nastygrams from Brian Ripley, so I'm feeling super confident.

The package is on Github here: https://github.com/PirateGrunt/raw_package. There are a few changes that I've not pushed yet, though I'll try to get to that tonight.

http://blog.revolutionanalytics.com/2015/12/microsoft-data-science.html
https://azure.microsoft.com/en-us/documentation/articles/machine-learning-data-science-provision-vm/

https://github.com/bescoto/RMSSQL
https://github.com/bescoto/dplyr.mssql

The above links contain short code examples. I can provide binary packages if anyone is interested, but I'm not sure if it's good practice to upload them to github.

Cheers,
Ben

I've had experience building csv file selection boxes in shiny online apps and have been intending to start this project for some time. But it would be more fun to work on this with other people — and might actually get done that way! Let me know if you're interested by replying to this post. Thanks.

Dan

http://blog.revolutionanalytics.com/2015/10/revolution-r-now-available-with-sql-server-community-preview.html

Glad to see Microsoft's buying Revolution R has some tangible benefits already. My company uses SQL Server, so I'm pretty excited about this. Does anyone know if there will be an extra fee, or if it's just included in SQL Server?

https://reserveprism.shinyapps.io/shiny_test

It seems to me the Shiny can achieve lots of web functions using R. They have done a lot of built-up to make it work. But it is pretty tricky if you don't know internet knowledge.

It is a very expensive server if you want to put it in production mode. And I don't know about security yet.

http://www.r-statistics.com/2015/06/setting-rstudio-server-using-amazon-web-services-aws-a-step-by-step-screenshots-tutorial/

http://community.wikidot.com/forum/t-1252893/cannot-delete-files

Similarly, would it be possible to have the RSS file allow links to be shown with an underline or a different colour? Right now it is hard to distinguish between text.

Thanks,
Simon

I just wanted to congratulate the group on the contributions to the blog so far. Most of them aren't in the draft blogs section yet, but we have a total of five submissions so far (from Greg McNulty, Simon Tam, Rajesh Sahasrabuddhe, Steve Yun, and myself). All of them (except possibly mine) seem to be on track to contributing useful knowledge on open source software to the community.

Once they start to show up on the draft blogs page we'll ask for reviewers. There will be quite a variety of topics, so if you're a perspective reviewer, hopefully one of these blog posts will interest you.

Thanks,
Ben

I'd like to maintain focus on the insurance applications for these tools. If anyone has prior experience with Hadoop and would like to collaborate on this piece, I'll happily share what I have thus far.

Thanks

Social hacking is a suggestion I made to the organizers. It's an informal environment where folks can bring a laptop, pose and answer questions and basically geek out for a bit. I've done similar things with data analysts here in the Triangle and it's quite a bit of fun. Something to bear in mind as a low stakes way to introduce people to R, Python, etc. at CAS events.

After expressing interest in contributing to the open source committee on the CAS volunteer survey I was contacted by Ben Escoto and said I would be willing to write a blog post about a project I did in R.

I work in reinsurance as a pricing actuary and we are always comparing experience and exposure for different excess layers. For casualty business, ALAE usually contributes a significant amount of excess layer losses. Most actuaries are probably familiar with the default assumption of ALAE = a fixed percentage of indemnity for all sizes of loss. After noticing how bad the exposure estimate was in layers heavily driven by ALAE (for example, a layer directly above a common primary policy limit)

I decided to do some research and found many publications on copula models for ALAE. My blogpost has a bibliography and all these sources are referenced there. The purpose of the blogpost is just to detail the R implementation I programmed and provide the code and auxiliary files. The model is certainly better and I hope the code provided can make it more practically accessible. I attached files to the blog, the code in a txt and all the upload and download csv files and some images.

If you're interested it needs a few additional reviewers. Thanks!

It was originally released as a FAViR paper. Should I do anything with this code? Should I try to get it on CRAN as a package?

Thanks for any thoughts,
Ben

I'm also going through the process of moving from .Rnw to .Rmd. This will support more output formats. The one I'm partial to at the moment is using the "Slidy" format for HTML5. Slidy is great because you can zoom in and out. I learned quickly that attendees had trouble reading the source code I'd put on a slide because the font was too small.

Feel free to fork the repository, submit pull requests, etc. It's a bit of a shambles at the moment, but should be clean fairly soon. Note that one of the branches is for a GitHub pages website with instructions and downloads for workshop attendees.

Repo may be found here: https://github.com/PirateGrunt/IntroToRforActuaries

My blog idea was to create a tutorial about my experience using some packages in R to write point and boundary data into a KML file. This file can then be viewed and further modified in Google Earth.

I used a combination of the tutorials found here in order to achieve the result.
Introduction to visualising spatial data in R (http://cran.r-project.org/doc/contrib/intro-spatial-rl.pdf)
plotKML package (http://gsif.isric.org/doku.php?id=wiki:tutorial_plotkml)

My hope is to have this ready by sometime in the new year for editing. Does anyone else have experience using these packages / tutorials?

Thanks,
Simon

setClass("Test", representation(changed="vector"),prototype(changed=c()))

setGeneric("getChanged", function(object, …) standardGeneric("getChanged"))
setMethod("getChanged", signature("Test"), function(object){return (object@changed)})

##################################################
#after this the changed is not changed at all, because it is inside an method,
#but even its onw method?
#################################################
setGeneric("test1", function(object, …) standardGeneric("test1"))
setMethod("test1", signature("Test"), function(object){object@changed<-c(1,2,3,4,5)})

###################################
#can this function change the value after call? Still not!!!
##################################
setGeneric("test2", function(object, …) standardGeneric("test2"))
setMethod("test2", signature("Test"), function(object){testing(object)<-c(1,2,3,4,5,6,7,8)})

setGeneric("test3", function(object, …) standardGeneric("test3"))
setMethod("test3", signature("Test"), function(object){testing(object)<-c(6,7,8, 9, 10); object})

#####################################
#replace a value for a S4 class, are we really replacing it?
#####################################
setGeneric("testing<-", function(this,value, …) standardGeneric("testing<-"))
setReplaceMethod("testing",signature("Test", "vector"), function(this, value) { this@changed<- value; this})

setMethod("print", signature("Test"), function(x){ print(getChanged(x))})

##########################################
#now, let us test it. PLEASE NOTE, this is in your Global environment!
#########################################
t<-new("Test", changed=c(1,2))

test1(t); print(t)
test2(t); print(t)
test3(t); print(t)
t<-test3(t); print(t)
testing(t)<-c("Hello","Open","souce", "committee"); print(t)

#but this is little wierd…
print(test1(t))
print(test2(t))
print(test3(t))
print(testing(t)<-c("done"))

Why Go?
This conference provides the chance to learn from some of the most knowledgeable people in the R community. The year’s line-up includes John Chambers, the creator of the precursor to R, S. In addition to increasing your R knowledge through the scheduled sessions, this conference creates an opportunity to discuss R with your fellow R users, see how other industries are using R, and engage in mutual learning.

My Personal Experience
I attended useR! at Vanderbilt University in 2012. Being fairly new to R, I was intimidated at first. But as I engaged others I found a vibrant community that was enthusiastic about sharing their knowledge and supporting each other. I also was able to ask questions and discuss R packages with the actual authors of the packages. One of the highlights was having a discussion at lunch with an internationally renowned statistician regarding the pros and cons of various tree based methods.

Location and Specifics
This year the conference is June 30 – July 3 at UCLA. Further information can be found at http://user2014.stat.ucla.edu/.

Can I say Chain Ladder looks like: professional guess from consecutive division average over summary, or something? If we use R for triangle analysis, what can we gain?

Ops, the comment shows under this blog again. Anyway, nice blog. Avi. But let me change the star level again. Interesting, the star level is averaged. Let me change it again.

Technically, open source software simply means that the source code for a piece of software is distributed together with, or in lieu of, its executable. It technically makes no statement about how the end user may take advantage of that source code, other than compiling it, and possibly learning from reading it.

The accepted definition of free software is software which is released under a license that guarantee users not only the freedom to run the software for any purpose, but also to study, modify, and distribute both the original software as well as any changed versions.

Because the word "free" has another connotation (Stallman's famous "free as in speech, not free as in beer" line) the industry has accepted the use of the term libre to mean "free as in beer" in this context.

So when people say free software, most of the time they are referring to FLOSS software: Free Libre Open Source Software. This is very different from freeware which means that the author is giving away the software, but not the source code or any license outside of using the software. For example, many iOS and Android apps are freeware, but if you reverse-engineer the source code and sell or give away your own version, you will quickly be sued for copyright infringement.

For anyone interested, there is some friction between the "free software" proponents and the "open source" software proponents. If you are interested, Wikipedia, the OSI, and the FSF are good places to look.

Even within the context of F(L)OSS, there are different classes of licenses which can be more or less restrictive. One of the most important differences between licenses is do derivative works need to be released under the exact same (or sometimes a similar enough) license, or may they be distributed however is desired, including on a proprietary basis (and for a fee). The licenses that require release under the same (or similar) licenses are called copyleft (yes, that is a pun on copyright). Examples include the Gnu GPL and LGPL. Other, more permissive, licenses include the BSD(-3) and Boost licenses.

Another important question is even if the released software will remain untouched, can it be "folded into" a proprietary piece of software or must any software into which it is compiled be release in toto on a free basis. This is the main difference between the GPL and LGPL. The former requires all software which includes any library licensed under the GPL to be released under a compatible license. The Gnu Lesser General Public License allows for a free library to be released in a proprietary package as long as the free library itself is untouched (and released with the proper license and author list etc.)

A third key question is with which licenses do we want our works to be compatible. For example, GPL2 is not compatible with the original BSD license (although GPL3 is compatible with BSD3) which meant that, legally, someone could not build a piece of software which relies on two libraries, one of which is GPL2 and the other OldBSD, as that would violate the license.

If we wanted to be as permissive as possible (allow someone to use our work to make money for his or her company, use our work to build proprietary software for sale, etc.) then we need to focus on licenses like the new BSD, the Boost, the ISC, the MIT, etc. which are pretty much along the lines "Use this for whatever you want at your own risk." If we wanted to ensure that our work remains open source we would have to focus on one of the more restrictive copyleft licenses.

Take some time, review the options, and let's have a discussion.

I wish we had an active blog, this would have been a good topic for a starting post :)

First order of business: Are you interested in joining? If so, please respond to this thread.

Second order of business: Let's change the name! I was thinking about Education and Outreach. Potential downside in that it abbreviates to E&O.

Third order of business: Discuss, share ideas and plan how to contribute to the CAS' efforts to increase the knowledge and use of R and other open source tools in our profession.

Here's what I'm up to:

Finalizing material for the R workshop at the RPM in DC.

Further development of my reserving package MRMR. I'm trying to nail down some basic mathematical operations and functions like subset, "[" etc. In addition, I'm trying to refine and extend the multidimensional treatment of triangle data.

Two weekends ago, I took part in Triangle Open Data Day. The Research Triangle in North Carolina is emerging as one of the leading areas of publicly accessible information. There are a number of groups working to provide apps that use public data for public good. When I can, I try to check in with the Durham Code for America brigade, who are doing some really cool stuff.

In the back of my head is the pressing need to finally get around to learning and using Python. Two months ago I went to a great presentation about scikit-learn, Python's machine learning library. This is enough of an impetus to see what all of the fuss is about.

One of these days I'll get back to my blog, where I hope to complete a post about waiting times for events that might not happen. We all know how to measure time until something DOES happen (death, claim closure, etc.), how about something that might not (whether you'll receive a promotion or a marriage proposal, whether you will observe the 59th claim when you've experienced 58)? There are 5 or 6 other blog posts that I need to get back to as well.

Finally, I'm thinking about May when I expect to be in Kigali, Rwanda talking about R. There's a project underway to instruct a host of folk there about survey methods, survival models and all manner of other techniques to investigate the spread of communicable disease.

And that's just a few weeks in the life of an open source actuary. What are YOU working on?

Dan Murphy

One thing I'd like to bring up here is developing a better understanding our target audience. For instance, if some actuaries are already using sophisticated methods and tools that are not open source, then I would venture to guess that they would not be our target audience. If other actuaries are currently using less sophisticated or efficient techniques, I'd be curious as to why this is the case. Is it due to budget constraints (i.e., can't afford the other software)? Is it because they have limited data? Is it because they don't even know how to use more advanced methods (i.e., an educational issue rather than a tool issue)?

I think once we know who the different possible customers are and understand their needs, then we can respond in a way that is most effective for them. We can set more specific goals for the committee, and better identify what resources we need to accomplish those goals (i.e., programmers, educators, how much time is needed, etc.).

My small pet-peeves are:

• Forums are hard to find (aren't in left navigation bar)
• URL links aren't underlined—-it's not clear what you can click on
• New pet peeve after previewing this message: the bullets in bulleted lists (like this one) don't show up

This is in addition to basic navigation / content issues, which could definitely use a broader conversation.

Thanks to all of you who filled out the survey (most of the committee did). Your responses contained some excellent suggestions for the coming year. I was also struck that the majority of you joined to learn and share knowledge of open source software. Most of you also committed to volunteering a significant amount of time for the committee.

Because this is a large committee (46 members), I suggest we organize our work by project, each with a clear goal that benefits the community. Please scan the projects below and think how you'd like to contribute. We'll also need someone to lead each project. I'll keep track of who's contributing to which project and will send out reminders to the people we don't hear from. Here's the list:

1. Wiki / Forums

We need one or more people to maintain this web site, and think how it might best serve the community. At the very least, it needs to provide a starting point for actuaries to learn about open source software that may be helpful to them. It might also be a good place for actuaries to learn about new relevant developments in the OSS world.

2. Webinars / Seminars / Presentations

In recent years there have been a number of R seminars, webinars, and presentations put on by actuaries, including by several members of this committee. The lead of this committee doesn't necessarily need to lead a seminar him/herself, but does need to keep in touch with the other relevant CAS committees and help coordinate webinars/seminars as required.

3. Actuarial R repository / Virtual Machine

Installing R and its various packages is a high barrier to entry for many actuaries. The hardware may not be available; IT policies may conflict; there is a changing list of packages to install. Lee Bowron and Dan Murphy made some progress last year with this issue—I believe the next step we discussed was to set up an Amazon Machine Image with everything pre-configured.

4. Blog

The majority of you said you joined in order to learn more about R and share what you know. I think the best format for this would be a high-quality blog. Contributors to this project would write-up their experiences with some application of open source software to an actuarial problem. We have a range of skills levels on the committee from beginner to advanced, so I'm sure a wide range of topics would be appreciated. The leader of this project would coordinate and schedule blog entries, and help figure out the implementation details.

5. Other

Don't feel limited by the above, please post below if you have other ideas or suggestions. For instance, recently committee members have mentioned the idea of an open-source textbook, or have thought about making R source code more popular for published articles. Keep in mind though, that to be successful projects need to produce some kind of beneficial, tangible output, and be realistic given the resources at our disposal.

Where to go from here

Please post in this thread about which project(s) you'd like to contribute to (or if you have a new idea). Mention too if you'd be willing to lead a particular project. In a couple weeks I'll take stock of the responses and see where we stand and if everyone's voiced an opinion.

In keeping with the open source spirit, this year I'd like us all to use the forum, instead of the email list. This will open up participation to people who may not be on the Open Source Committee mailing list (or who may not be associated with the CAS at all). Offhand I can't think of much forum business that couldn't be done in the open on this forum, but use your best judgment of course.

This forum supports RSS, so you can get automatic notifications through Outlook, Gmail, or another RSS reader. All committee members should be subscribed to this "Committee Business" forum. The RSS link is:

http://opensourcesoftware.casact.org/feed/forum/cp-115920.xml

You'll probably also want to subscribe to the "Open Source" forum at:

http://opensourcesoftware.casact.org/feed/forum/cp-115921.xml

Here are instructions for using Outlook 2007 as an RSS reader. If you're like me you only check your inbox, so you may want to move the detail folder to your Inbox. Here is a recent article on Gmail (I don't use Gmail, so I haven't tried this one myself).

Please post below if you have any comments/concerns/suggestions about the forums.

Thanks,
Ben Escoto

On Tuesday we discussed different ways of connecting R and Excel. Here are some links to the Excel add-in I've been working on, Excelsi-R:

• Excelsi-R home page
• Files for download
• User documentation

The Excel add-in lets you run R commands from Excel formulas or put R graphs in Excel spreadsheets. The R package contains some support functions for the add-in as well as functions for reading and writing to Excel or the Windows clipboard.

It's still in alpha but the core functionality should work. I'd appreciate any comments or suggestions.

"The purpose of the Open-Source Software Committee is to make advanced open-source reserving, ratemaking and predictive modeling procedures available, accessible and of practical value to a wider audience. The Open-Source Software Research Committee will foster the development of tutorials, interfaces, procedures, and educational resources, and will liaise with the Professional Education committees to disseminate information to CAS members."

What goals should we make in 2013?

For continuing education, what topics are most relevant?

What needs to be added to the wiki?

Any other ideas?

Thanks, Lee

http://opensourcesoftware.casact.org/nav:top

2. Click "edit" at the bottom

3. Should look something like this:

* [# Articles] * [[[Useful Resources]]] * [[[Introduction to R]]] * [[[Text Mining]]] * [[[Chain Ladder]]] * [[[The title of your article here]]]

[[div style="text-indent: 0px;"]] {{code here}} [[/div]]

the div statement in brackets removes the automatic indentation and the braces make the text teletyped. really makes the code stick out from normal text. examples can be seen in the chain ladder article.

[[div class="links-blue"]]

[[/div]

Changing the color in the code doesn't do anything except make the code not work. The color will change for any link in text contained in the preceding brackets; you do not have to put the code at the beginning and end, it's just a simple method to make sure all links are affected.

1) About transferring from Excel to/from R: I think saving as csv and using read.table is a pretty cumbersome way of interacting with Excel and will put people off, especially people who usually only use Excel. Personally, I have some simple CopyClip() and Paste() functions which can be used to copy and paste data frames to and from Excel. If anyone is interested I can post them, but there may be better stuff out there. Also, there is RExcel, which is more sophisticated, but I could never get it installed.

2) About "Data Storage in R": I think it gives a misleading impression about R and memory. In almost all mainstream computer languages with the exception of SQL, any created object is by default created in memory. This is true of C, C#, java, python, VBA, whatever. No one complains that C is memory limited, but R's treatment of memory is the same. Furthermore, R has great packages like RODBC and especially sqldf which make it very easy to move data into and out of databases. Whether or not R runs only in memory depends only on the implementation of the particular R function. For instance, the default lm function loads all of its data into memory, but the biglm function leaves most of its data on disk. I think you could just say something like "R is similar to most mainstream computer languages—-most of its functions operate on objects in system memory. Functions must be specifically written to use disk space instead of RAM. However, R includes several packages that facilitate reading and writing to popular databases.

3) Is the sound supposed to work in the AVI files? I didn't hear anything, but perhaps it was an issue with my computer.

We need to decide what the output of our subgroup will be. Lee Bowron
suggested the following as a rough outline:

1. Identifying suitable reserving data sets
2. Reviewing reserving methods available in R.
3. Using at least one method to analyze a data set.

For (1) above, here are some basic triangle reserving data sets that
I'm aware of:

- Jed Frees's R data includes several triangle sets
- The Chainladder package has some reserving data
- We could always input annual statement data or type in sample data
from a published paper.

About (2), here are some methods already implemented and freely
available in R:

- Mack, Bootstrap, and Munich Chain Ladder in the Chainladder package
- Glenn Meyers' Collective Risk Model
- Frank Schmid and Chris Laws' lossDev method

About (3), it seems we have to choose which method to employ and how
to package our work. As to the method, I think there is a tradeoff
between being sophisticated and showing off what R can do (e.g. the
lossDev package) and doing something simpler that all actuaries are
familiar with and can understand (e.g. Chain Ladder).

My current inclination is to rework the FAViR basic reserving paper,
and present the Chain Ladder method (either Mack or Bootstrap) and the
Additive Loss Method (i.e. the Cape Cod Method). Actuaries will be
familiar with (at least with the non-stochastic components of) these,
and we can show how R can present the old LDF vs BF battle in a
regression context to analyze fit and provide reserve ranges. It
seems if we use a more sophisticated method, actuaries would want to
see how the results compare to a more traditional chainladder and BF
method.

We can pick a data set to use as an example, but creating an Excel
template (like the Parallelogram example I sent to all of you) would
allow actuaries to plug in their own data and get rolling with these
methods without having to type any source code or install anything.
Hopefully actuaries will see that it works and makes nice graphs, and
will then be motivated to install and start using R.

Any thoughts? Does anyone know of any more data sets or reserving methods available in R? How do you feel about the basic plan above?

Links for data/methods mentioned above:

Jed Frees's data files, including triangle data:
http://research3.bus.wisc.edu/file.php/129/RTSABook/WebData/RegressionTSProjectDataFiles.htm

Chainladder package and reserving data:
http://code.google.com/p/chainladder/

lossDev package:
http://lossdev.r-forge.r-project.org/

Glenn Meyers' collective risk model paper with code:
http://www.variancejournal.org/issues/?fa=article&abstrID=6606n

Example FAViR template:
http://www.favir.net/excel

FAViR "Basic Reserving" paper:
http://www.favir.net/papers

Benedict Escoto (Lead)
Sholom Feldblum
Thomas Kolde
Scott Sobel

However, I thought it might be better for us to use these new forums instead of email. With the forums, anyone can read these messages or comment, whether or not they were assigned to the reserving subgroup. Now I'll repost to this forum a couple of emails to catch everyone up.