Re: [julia-users] Re: documentation suggestions

[Po Choi]

Before seriously hiring someone to bring Julia into your school, perhaps 
you first can try the commercial service from http://juliacomputing.com/ to 
organize some workshops or events to see how the students and other 
faculties feel about the potential of Julia.

On Wednesday, February 10, 2016 at 3:26:34 PM UTC-8, ivo welch wrote:
>
>
> ooops...I leaked my signature.   not a problem, but it is also was not 
> necessarily what I had meant to say.  for those who are interested, here is 
> a little background from my side of the world.
>
> ucla anderson, like most other business schools, has been pretty ignorant 
> with respect to any kind of research computing expertise.
>
> this is beginning to change, as management schools (incl us) are moving 
> towards one-year quantitatively oriented one-year masters program. 
>  anderson already has a masters of financial engineering and is about to 
> start a masters program in data analytics.  as for me, I am also trying to 
> figure out how to offer more of this to MBA students, our traditional bread 
> and butter, but it is not clear whether this can be implemented.  so, in 
> the future, we will need more data, programming, and other computing 
> support than we did in the past.  like every other industry.
>
> it is exceedingly difficult to hire good programmers in a context like 
> our's.  universities do not pay much, for institutional reasons. 
>  individuals that are very good at this tend to be lured away to industry 
> if they are good, and non-terminable if they are bad.  a year goes by very 
> fast---we may find someone for one year, but then not the next.  any 
> program has to be prepared to run for decades.  we cannot shut down a 
> masters program for lack of a critical person.
>
> our current IT department (both UCLA and Anderson) mostly handle basics, 
> such as the network and Microsoft apps.  as far as I can tell, 
> http://www.ats.ucla.edu/stat/ offers some R expertise, but not Julia 
> expertise.  its depth has varied with the individuals working there.  there 
> is no julia support afaik.
>
> our best choices are typically individuals that want to get a phd and just 
> happen to have good expertise.  R, julia, etc.  another choice would be 
> someone who wants to work half-time on a project like julia and the other 
> half-time work on direct program support.  job has nice benefits...
>
> just to get a position approved can take UC about 3-6 months and is a 
> high-effort affair.  we have rules up the wazoo.  there is also one month 
> of data expertise that anyone would want to learn (WRDS, CRSP, Compustat).  
> I can spend a month full-time to get there.  sigh.
>
> so, for the most part, the few of us faculty and phd students, who like 
> programming have been bootstrapping it ourselves.  at UCLA Anderson, we are 
> luckier in this respect than many other places (Keith Chen, Peter Rossi, 
> John Mamer, ...), but it's tough.
>
> julia expertise would be great for us to have.  it would have great 
> externalities for us.  if anyone with deep julia expertise wants to apply 
> to UCLA for a few years (phd, undergrad, master), with a side job at 
> Anderson, then drop me an email ;-).  for obvious reasons, faculty has and 
> wants no power to make admission decisions (or we would be besieged by our 
> friends and family), but I could put in a good word with our admissions 
> department(s).  it matters on the margin.  if someone working on julia 
> wants a regular job, also please email me.
>
> /iaw
> 
> Ivo Welch (ivo@gmail.com )
> http://www.ivo-welch.info/
>
>
> On Wed, Feb 10, 2016 at 2:47 PM, Jeffrey Sarnoff  > wrote:
>
>> That is a reasonable want; it may take Anderson some time to institute 
>> scholarships for expertise in Julia
>> If you were already expert with Julia, what would you have your students 
>> doing?
>>
>>
>>   for expertThat is a reasonable want.  As an alternative, Anderson is 
>> not offering scholarships earmarked for Julia experts. 
>>
>> On Wednesday, February 10, 2016 at 3:49:47 PM UTC-5, ivo welch wrote:
>>>
>>>
>>> indeed.  thank you, josh.  I would add a final chapter at 
>>>
>>> http://docs.julialang.org/en/release-0.4/
>>>
>>> with a set of links to various further resources, examples, full 
>>> stand-alone programs, etc.  for me, at least, the perl cookbook and sets of 
>>> self-contained snippet programs to start with, were the main reason why I 
>>> learned perl many years ago.
>>>
>>> the key problem to my use of julia over R for my students is that I do 
>>> not have a resident julia expert at UCLA.  this won't change anytime soon, 
>>> because they are hard to find (hire) :-(.  this google forum is great, but 
>>> it's scary to switch without a double hull.  many, many full *working* 
>>> standalone examples are the next best thing for me.
>>>
>>> regards,
>>>
>>> /iaw
>>>
>>>
>>> 
>>> Ivo Welch (ivo@gmail.com)
>>> http://www.ivo-welch.info/
>>> J. Fred Weston Distinguished Professor of Fina

Re: [julia-users] Re: documentation suggestions

[Kristoffer Carlsson]

It is hard to actually lose work with git. With git reflog you can always see 
where your HEAD has been and you can then git reset to a previous revision. 

Re: [julia-users] Re: documentation suggestions

[Jeffrey Sarnoff]

Doug,  I found some shelter after months of being bit by git using this 
(free for non-commercial use) www.syntevo.com/smartgit/ 



On Friday, February 12, 2016 at 12:49:27 PM UTC-5, Douglas Bates wrote:
>
> On Thursday, February 11, 2016 at 3:06:45 PM UTC-6, ivo welch wrote:
>>
>>
>> hi doug---and vice-versa.  it's interesting that a core function (reading 
>> a .csv file) would not be in a native julia library.  when are you 
>> switching your students to julia?  regards,  /iaw
>>
>
> Writing a function to read a .csv file is not trivial - partly because CSV 
> is not well-defined.  It is also the case of an itch getting scratched - if 
> those working on Julia with the skills to write such a function don't have 
> a need to read .csv files that particular functionality stagnates.
>
> The definition and functionality of data frames, which are the natural 
> output when reading a CSV file,  in Julia is still being debated.  In R the 
> choices were much easier because R was designed to emulate S version 3 in 
> which a data frame was a central construct.  Sacrifices in performance were 
> made to allow for checking for NA's during each atomic arithmetic 
> operation.  That trade-off wouldn't fly in Julia.  Also R vector structures 
> all allow for element names - again at an expense in performance.
>
> I'm not really in the position to convert my students as I am now an 
> Emeritus Professor.  I do still offer a seminar series on "Statistics with 
> Julia" and have convinced some students to use Julia in thesis research.
>
> I would be quite happy with Julia if only git and I got along better.  I 
> just lost three days worth of work this morning because of yet another git 
> disaster.
>
>>
>> 
>> Ivo Welch (ivo@gmail.com)
>> http://www.ivo-welch.info/
>> J. Fred Weston Distinguished Professor of Finance
>> Anderson School at UCLA, C519
>> Free Finance Textbook, http://book.ivo-welch.info/
>> Exec Editor, Critical Finance Review, 
>> http://www.critical-finance-review.org/
>> Editor and Publisher, FAMe, http://www.fame-jagazine.com/
>>
>> On Thu, Feb 11, 2016 at 12:37 PM, Douglas Bates  wrote:
>>
>>> Hi Ivo,
>>>
>>> Good to hear from you.
>>>
>>> On Wednesday, February 10, 2016 at 9:58:37 AM UTC-6, ivo welch wrote:


 ladies and gents---I am not (yet) a julia user.

 may I suggest adding more examples into two places where julia users 
 will face starting hurdles?

 [1] the I/O docs of julia.  like, reading and writing csv files that 
 are compressed and decompressed on-the-fly, even if not in the ultimate 
 efficient manner.a large fraction of the time and frustration of new 
 users is consumed by the task of shoehorning data into and out of new 
 computer languages.  with all of R's problem, the ' d <- 
 read.csv("f.csv")' 
 and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry 
 frustration greatly.  perhaps xml file reading and writing.  perhaps...

 [2] more 'standard task' programs would be great.  read a csv file, run 
 a regression according to variable names on the command line, print 
 output, 
 draw a graph.  I know there are fragments throughout the docs, but some 
 section with ready to run complete programs would be good, perhaps at the 
 end of the manual.

 in a year, I hope to switch my students from R to julia.

>>>
>>> My main use of the RCall package is to import datasets from R into 
>>> Julia.  If I have a dataset in an R package I use, e.g.
>>>
>>>  julia> using RCall
>>>
>>> julia> ds = rcopy("lme4::Dyestuff")
>>> 30x2 DataFrames.DataFrame
>>> | Row | Batch | Yield  |
>>> |-|---||
>>> | 1   | "A"   | 1545.0 |
>>> | 2   | "A"   | 1440.0 |
>>> | 3   | "A"   | 1440.0 |
>>> | 4   | "A"   | 1520.0 |
>>> | 5   | "A"   | 1580.0 |
>>> | 6   | "B"   | 1540.0 |
>>> | 7   | "B"   | 1555.0 |
>>> | 8   | "B"   | 1490.0 |
>>> | 9   | "B"   | 1560.0 |
>>> | 10  | "B"   | 1495.0 |
>>> | 11  | "C"   | 1595.0 |
>>> | 12  | "C"   | 1550.0 |
>>> | 13  | "C"   | 1605.0 |
>>> | 14  | "C"   | 1510.0 |
>>> | 15  | "C"   | 1560.0 |
>>> | 16  | "D"   | 1445.0 |
>>> | 17  | "D"   | 1440.0 |
>>> | 18  | "D"   | 1595.0 |
>>> | 19  | "D"   | 1465.0 |
>>> | 20  | "D"   | 1545.0 |
>>> | 21  | "E"   | 1595.0 |
>>> | 22  | "E"   | 1630.0 |
>>> | 23  | "E"   | 1515.0 |
>>> | 24  | "E"   | 1635.0 |
>>> | 25  | "E"   | 1625.0 |
>>> | 26  | "F"   | 1520.0 |
>>> | 27  | "F"   | 1455.0 |
>>> | 28  | "F"   | 1450.0 |
>>> | 29  | "F"   | 1480.0 |
>>> | 30  | "F"   | 1445.0 |
>>>
>>> If I wanted to read a CSV file using the facilities in R I could use
>>>
>>> julia> rcopy("read.csv('/usr/share/distro-info/debian.csv')")
>>> 17x6 DataFrames.DataFrame
>>> | Row | version | codename   | series | created  | 
>>> release  | eol  |
>>>
>>> |-|-|||--|--|--|
>>> | 1   | 1.1 | "Buzz"

Re: [julia-users] Re: documentation suggestions

[Douglas Bates]

On Thursday, February 11, 2016 at 3:06:45 PM UTC-6, ivo welch wrote:
>
>
> hi doug---and vice-versa.  it's interesting that a core function (reading 
> a .csv file) would not be in a native julia library.  when are you 
> switching your students to julia?  regards,  /iaw
>

Writing a function to read a .csv file is not trivial - partly because CSV 
is not well-defined.  It is also the case of an itch getting scratched - if 
those working on Julia with the skills to write such a function don't have 
a need to read .csv files that particular functionality stagnates.

The definition and functionality of data frames, which are the natural 
output when reading a CSV file,  in Julia is still being debated.  In R the 
choices were much easier because R was designed to emulate S version 3 in 
which a data frame was a central construct.  Sacrifices in performance were 
made to allow for checking for NA's during each atomic arithmetic 
operation.  That trade-off wouldn't fly in Julia.  Also R vector structures 
all allow for element names - again at an expense in performance.

I'm not really in the position to convert my students as I am now an 
Emeritus Professor.  I do still offer a seminar series on "Statistics with 
Julia" and have convinced some students to use Julia in thesis research.

I would be quite happy with Julia if only git and I got along better.  I 
just lost three days worth of work this morning because of yet another git 
disaster.

>
> 
> Ivo Welch (ivo@gmail.com )
> http://www.ivo-welch.info/
> J. Fred Weston Distinguished Professor of Finance
> Anderson School at UCLA, C519
> Free Finance Textbook, http://book.ivo-welch.info/
> Exec Editor, Critical Finance Review, 
> http://www.critical-finance-review.org/
> Editor and Publisher, FAMe, http://www.fame-jagazine.com/
>
> On Thu, Feb 11, 2016 at 12:37 PM, Douglas Bates  > wrote:
>
>> Hi Ivo,
>>
>> Good to hear from you.
>>
>> On Wednesday, February 10, 2016 at 9:58:37 AM UTC-6, ivo welch wrote:
>>>
>>>
>>> ladies and gents---I am not (yet) a julia user.
>>>
>>> may I suggest adding more examples into two places where julia users 
>>> will face starting hurdles?
>>>
>>> [1] the I/O docs of julia.  like, reading and writing csv files that are 
>>> compressed and decompressed on-the-fly, even if not in the ultimate 
>>> efficient manner.a large fraction of the time and frustration of new 
>>> users is consumed by the task of shoehorning data into and out of new 
>>> computer languages.  with all of R's problem, the ' d <- read.csv("f.csv")' 
>>> and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry 
>>> frustration greatly.  perhaps xml file reading and writing.  perhaps...
>>>
>>> [2] more 'standard task' programs would be great.  read a csv file, run 
>>> a regression according to variable names on the command line, print output, 
>>> draw a graph.  I know there are fragments throughout the docs, but some 
>>> section with ready to run complete programs would be good, perhaps at the 
>>> end of the manual.
>>>
>>> in a year, I hope to switch my students from R to julia.
>>>
>>
>> My main use of the RCall package is to import datasets from R into 
>> Julia.  If I have a dataset in an R package I use, e.g.
>>
>>  julia> using RCall
>>
>> julia> ds = rcopy("lme4::Dyestuff")
>> 30x2 DataFrames.DataFrame
>> | Row | Batch | Yield  |
>> |-|---||
>> | 1   | "A"   | 1545.0 |
>> | 2   | "A"   | 1440.0 |
>> | 3   | "A"   | 1440.0 |
>> | 4   | "A"   | 1520.0 |
>> | 5   | "A"   | 1580.0 |
>> | 6   | "B"   | 1540.0 |
>> | 7   | "B"   | 1555.0 |
>> | 8   | "B"   | 1490.0 |
>> | 9   | "B"   | 1560.0 |
>> | 10  | "B"   | 1495.0 |
>> | 11  | "C"   | 1595.0 |
>> | 12  | "C"   | 1550.0 |
>> | 13  | "C"   | 1605.0 |
>> | 14  | "C"   | 1510.0 |
>> | 15  | "C"   | 1560.0 |
>> | 16  | "D"   | 1445.0 |
>> | 17  | "D"   | 1440.0 |
>> | 18  | "D"   | 1595.0 |
>> | 19  | "D"   | 1465.0 |
>> | 20  | "D"   | 1545.0 |
>> | 21  | "E"   | 1595.0 |
>> | 22  | "E"   | 1630.0 |
>> | 23  | "E"   | 1515.0 |
>> | 24  | "E"   | 1635.0 |
>> | 25  | "E"   | 1625.0 |
>> | 26  | "F"   | 1520.0 |
>> | 27  | "F"   | 1455.0 |
>> | 28  | "F"   | 1450.0 |
>> | 29  | "F"   | 1480.0 |
>> | 30  | "F"   | 1445.0 |
>>
>> If I wanted to read a CSV file using the facilities in R I could use
>>
>> julia> rcopy("read.csv('/usr/share/distro-info/debian.csv')")
>> 17x6 DataFrames.DataFrame
>> | Row | version | codename   | series | created  | 
>> release  | eol  |
>>
>> |-|-|||--|--|--|
>> | 1   | 1.1 | "Buzz" | "buzz" | "1993-08-16" | 
>> "1996-06-17" | "1997-06-05" |
>> | 2   | 1.2 | "Rex"  | "rex"  | "1996-06-17" | 
>> "1996-12-12" | "1998-06-05" |
>> | 3   | 1.3 | "Bo"   | "bo"   | "1996-12-12" | 
>> "1997-06-05" | "1999-03-09" |
>> | 4   | 2.0 | "Hamm" | "hamm" | "1997-06-05" |

Re: [julia-users] Re: documentation suggestions

[Ariel Katz]

Hello,

With regards to your specific point about CSV I/O,   there are a several 
ways to read CSV files in Julia.

- Dataframes 

.jl:

df = readtable("data.csv")


- Base:

readdlm(source, delim::Char, T::Type; options...)

- And the current state of the art  with regards to speed, CSV.jl 
 with its datastream integration.

Unless you are reading fairly large CSV files, I would stick to 
Dataframes.jl.

I would caution you though that Data I/O in Julia is still in its infancy 
and there are methods that are either slower than Python/R or missing (xls 
etc).

Zooming out a bit, I've found Data Wookie Month of Julia 
 blog series to be the best 
getting started guide for practical data sciency Julia stuff. 

On Thursday, February 11, 2016 at 4:06:45 PM UTC-5, ivo welch wrote:
>
>
> hi doug---and vice-versa.  it's interesting that a core function (reading 
> a .csv file) would not be in a native julia library.  when are you 
> switching your students to julia?  regards,  /iaw
>
>
> 
> Ivo Welch (ivo@gmail.com )
> http://www.ivo-welch.info/
> J. Fred Weston Distinguished Professor of Finance
> Anderson School at UCLA, C519
> Free Finance Textbook, http://book.ivo-welch.info/
> Exec Editor, Critical Finance Review, 
> http://www.critical-finance-review.org/
> Editor and Publisher, FAMe, http://www.fame-jagazine.com/
>
> On Thu, Feb 11, 2016 at 12:37 PM, Douglas Bates  > wrote:
>
>> Hi Ivo,
>>
>> Good to hear from you.
>>
>> On Wednesday, February 10, 2016 at 9:58:37 AM UTC-6, ivo welch wrote:
>>>
>>>
>>> ladies and gents---I am not (yet) a julia user.
>>>
>>> may I suggest adding more examples into two places where julia users 
>>> will face starting hurdles?
>>>
>>> [1] the I/O docs of julia.  like, reading and writing csv files that are 
>>> compressed and decompressed on-the-fly, even if not in the ultimate 
>>> efficient manner.a large fraction of the time and frustration of new 
>>> users is consumed by the task of shoehorning data into and out of new 
>>> computer languages.  with all of R's problem, the ' d <- read.csv("f.csv")' 
>>> and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry 
>>> frustration greatly.  perhaps xml file reading and writing.  perhaps...
>>>
>>> [2] more 'standard task' programs would be great.  read a csv file, run 
>>> a regression according to variable names on the command line, print output, 
>>> draw a graph.  I know there are fragments throughout the docs, but some 
>>> section with ready to run complete programs would be good, perhaps at the 
>>> end of the manual.
>>>
>>> in a year, I hope to switch my students from R to julia.
>>>
>>
>> My main use of the RCall package is to import datasets from R into 
>> Julia.  If I have a dataset in an R package I use, e.g.
>>
>>  julia> using RCall
>>
>> julia> ds = rcopy("lme4::Dyestuff")
>> 30x2 DataFrames.DataFrame
>> | Row | Batch | Yield  |
>> |-|---||
>> | 1   | "A"   | 1545.0 |
>> | 2   | "A"   | 1440.0 |
>> | 3   | "A"   | 1440.0 |
>> | 4   | "A"   | 1520.0 |
>> | 5   | "A"   | 1580.0 |
>> | 6   | "B"   | 1540.0 |
>> | 7   | "B"   | 1555.0 |
>> | 8   | "B"   | 1490.0 |
>> | 9   | "B"   | 1560.0 |
>> | 10  | "B"   | 1495.0 |
>> | 11  | "C"   | 1595.0 |
>> | 12  | "C"   | 1550.0 |
>> | 13  | "C"   | 1605.0 |
>> | 14  | "C"   | 1510.0 |
>> | 15  | "C"   | 1560.0 |
>> | 16  | "D"   | 1445.0 |
>> | 17  | "D"   | 1440.0 |
>> | 18  | "D"   | 1595.0 |
>> | 19  | "D"   | 1465.0 |
>> | 20  | "D"   | 1545.0 |
>> | 21  | "E"   | 1595.0 |
>> | 22  | "E"   | 1630.0 |
>> | 23  | "E"   | 1515.0 |
>> | 24  | "E"   | 1635.0 |
>> | 25  | "E"   | 1625.0 |
>> | 26  | "F"   | 1520.0 |
>> | 27  | "F"   | 1455.0 |
>> | 28  | "F"   | 1450.0 |
>> | 29  | "F"   | 1480.0 |
>> | 30  | "F"   | 1445.0 |
>>
>> If I wanted to read a CSV file using the facilities in R I could use
>>
>> julia> rcopy("read.csv('/usr/share/distro-info/debian.csv')")
>> 17x6 DataFrames.DataFrame
>> | Row | version | codename   | series | created  | 
>> release  | eol  |
>>
>> |-|-|||--|--|--|
>> | 1   | 1.1 | "Buzz" | "buzz" | "1993-08-16" | 
>> "1996-06-17" | "1997-06-05" |
>> | 2   | 1.2 | "Rex"  | "rex"  | "1996-06-17" | 
>> "1996-12-12" | "1998-06-05" |
>> | 3   | 1.3 | "Bo"   | "bo"   | "1996-12-12" | 
>> "1997-06-05" | "1999-03-09" |
>> | 4   | 2.0 | "Hamm" | "hamm" | "1997-06-05" | 
>> "1998-07-24" | "2000-03-09" |
>> | 5   | 2.1 | "Slink"| "slink"| "1998-07-24" | 
>> "1999-03-09" | "2000-10-30" |
>> | 6   | 2.2 | "Potato"   | "potato"   | "1999-03-09" | 
>> "2000-08-15" | "2003-07-30" |
>> | 7   | 3.0 | "Woody"| "woo

Re: [julia-users] Re: documentation suggestions

[ivo welch]

hi doug---and vice-versa.  it's interesting that a core function (reading a
.csv file) would not be in a native julia library.  when are you switching
your students to julia?  regards,  /iaw



Ivo Welch (ivo.we...@gmail.com)
http://www.ivo-welch.info/
J. Fred Weston Distinguished Professor of Finance
Anderson School at UCLA, C519
Free Finance Textbook, http://book.ivo-welch.info/
Exec Editor, Critical Finance Review,
http://www.critical-finance-review.org/
Editor and Publisher, FAMe, http://www.fame-jagazine.com/

On Thu, Feb 11, 2016 at 12:37 PM, Douglas Bates  wrote:

> Hi Ivo,
>
> Good to hear from you.
>
> On Wednesday, February 10, 2016 at 9:58:37 AM UTC-6, ivo welch wrote:
>>
>>
>> ladies and gents---I am not (yet) a julia user.
>>
>> may I suggest adding more examples into two places where julia users will
>> face starting hurdles?
>>
>> [1] the I/O docs of julia.  like, reading and writing csv files that are
>> compressed and decompressed on-the-fly, even if not in the ultimate
>> efficient manner.a large fraction of the time and frustration of new
>> users is consumed by the task of shoehorning data into and out of new
>> computer languages.  with all of R's problem, the ' d <- read.csv("f.csv")'
>> and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry
>> frustration greatly.  perhaps xml file reading and writing.  perhaps...
>>
>> [2] more 'standard task' programs would be great.  read a csv file, run a
>> regression according to variable names on the command line, print output,
>> draw a graph.  I know there are fragments throughout the docs, but some
>> section with ready to run complete programs would be good, perhaps at the
>> end of the manual.
>>
>> in a year, I hope to switch my students from R to julia.
>>
>
> My main use of the RCall package is to import datasets from R into Julia.
> If I have a dataset in an R package I use, e.g.
>
>  julia> using RCall
>
> julia> ds = rcopy("lme4::Dyestuff")
> 30x2 DataFrames.DataFrame
> | Row | Batch | Yield  |
> |-|---||
> | 1   | "A"   | 1545.0 |
> | 2   | "A"   | 1440.0 |
> | 3   | "A"   | 1440.0 |
> | 4   | "A"   | 1520.0 |
> | 5   | "A"   | 1580.0 |
> | 6   | "B"   | 1540.0 |
> | 7   | "B"   | 1555.0 |
> | 8   | "B"   | 1490.0 |
> | 9   | "B"   | 1560.0 |
> | 10  | "B"   | 1495.0 |
> | 11  | "C"   | 1595.0 |
> | 12  | "C"   | 1550.0 |
> | 13  | "C"   | 1605.0 |
> | 14  | "C"   | 1510.0 |
> | 15  | "C"   | 1560.0 |
> | 16  | "D"   | 1445.0 |
> | 17  | "D"   | 1440.0 |
> | 18  | "D"   | 1595.0 |
> | 19  | "D"   | 1465.0 |
> | 20  | "D"   | 1545.0 |
> | 21  | "E"   | 1595.0 |
> | 22  | "E"   | 1630.0 |
> | 23  | "E"   | 1515.0 |
> | 24  | "E"   | 1635.0 |
> | 25  | "E"   | 1625.0 |
> | 26  | "F"   | 1520.0 |
> | 27  | "F"   | 1455.0 |
> | 28  | "F"   | 1450.0 |
> | 29  | "F"   | 1480.0 |
> | 30  | "F"   | 1445.0 |
>
> If I wanted to read a CSV file using the facilities in R I could use
>
> julia> rcopy("read.csv('/usr/share/distro-info/debian.csv')")
> 17x6 DataFrames.DataFrame
> | Row | version | codename   | series | created  | release
>  | eol  |
>
> |-|-|||--|--|--|
> | 1   | 1.1 | "Buzz" | "buzz" | "1993-08-16" |
> "1996-06-17" | "1997-06-05" |
> | 2   | 1.2 | "Rex"  | "rex"  | "1996-06-17" |
> "1996-12-12" | "1998-06-05" |
> | 3   | 1.3 | "Bo"   | "bo"   | "1996-12-12" |
> "1997-06-05" | "1999-03-09" |
> | 4   | 2.0 | "Hamm" | "hamm" | "1997-06-05" |
> "1998-07-24" | "2000-03-09" |
> | 5   | 2.1 | "Slink"| "slink"| "1998-07-24" |
> "1999-03-09" | "2000-10-30" |
> | 6   | 2.2 | "Potato"   | "potato"   | "1999-03-09" |
> "2000-08-15" | "2003-07-30" |
> | 7   | 3.0 | "Woody"| "woody"| "2000-08-15" |
> "2002-07-19" | "2006-06-30" |
> | 8   | 3.1 | "Sarge"| "sarge"| "2002-07-19" |
> "2005-06-06" | "2008-03-30" |
> | 9   | 4.0 | "Etch" | "etch" | "2005-06-06" |
> "2007-04-08" | "2010-02-15" |
> | 10  | 5.0 | "Lenny"| "lenny"| "2007-04-08" |
> "2009-02-14" | "2012-02-06" |
> | 11  | 6.0 | "Squeeze"  | "squeeze"  | "2009-02-14" |
> "2011-02-06" | "2014-05-31" |
> | 12  | 7.0 | "Wheezy"   | "wheezy"   | "2011-02-06" |
> "2013-05-04" | ""   |
> | 13  | 8.0 | "Jessie"   | "jessie"   | "2013-05-04" |
> "2015-04-25" | ""   |
> | 14  | 9.0 | "Stretch"  | "stretch"  | "2015-04-25" | ""
>   | ""   |
> | 15  | 10.0| "Buster"   | "buster"   | "2018-07-01" | ""
>   | ""   |
> | 16  | NA  | "Sid"  | "sid"  | "1993-08-16" | ""
>   | ""   |
> | 17  | NA  | "Experimental" | "experimental" | "1993-08-16" | ""
>   | ""   |
>
>
> (It turns out that R's allowing either ' or " for enclosing str

[julia-users] Re: documentation suggestions

[Douglas Bates]

Hi Ivo,

Good to hear from you.

On Wednesday, February 10, 2016 at 9:58:37 AM UTC-6, ivo welch wrote:
>
>
> ladies and gents---I am not (yet) a julia user.
>
> may I suggest adding more examples into two places where julia users will 
> face starting hurdles?
>
> [1] the I/O docs of julia.  like, reading and writing csv files that are 
> compressed and decompressed on-the-fly, even if not in the ultimate 
> efficient manner.a large fraction of the time and frustration of new 
> users is consumed by the task of shoehorning data into and out of new 
> computer languages.  with all of R's problem, the ' d <- read.csv("f.csv")' 
> and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry 
> frustration greatly.  perhaps xml file reading and writing.  perhaps...
>
> [2] more 'standard task' programs would be great.  read a csv file, run a 
> regression according to variable names on the command line, print output, 
> draw a graph.  I know there are fragments throughout the docs, but some 
> section with ready to run complete programs would be good, perhaps at the 
> end of the manual.
>
> in a year, I hope to switch my students from R to julia.
>

My main use of the RCall package is to import datasets from R into Julia. 
 If I have a dataset in an R package I use, e.g.

 julia> using RCall

julia> ds = rcopy("lme4::Dyestuff")
30x2 DataFrames.DataFrame
| Row | Batch | Yield  |
|-|---||
| 1   | "A"   | 1545.0 |
| 2   | "A"   | 1440.0 |
| 3   | "A"   | 1440.0 |
| 4   | "A"   | 1520.0 |
| 5   | "A"   | 1580.0 |
| 6   | "B"   | 1540.0 |
| 7   | "B"   | 1555.0 |
| 8   | "B"   | 1490.0 |
| 9   | "B"   | 1560.0 |
| 10  | "B"   | 1495.0 |
| 11  | "C"   | 1595.0 |
| 12  | "C"   | 1550.0 |
| 13  | "C"   | 1605.0 |
| 14  | "C"   | 1510.0 |
| 15  | "C"   | 1560.0 |
| 16  | "D"   | 1445.0 |
| 17  | "D"   | 1440.0 |
| 18  | "D"   | 1595.0 |
| 19  | "D"   | 1465.0 |
| 20  | "D"   | 1545.0 |
| 21  | "E"   | 1595.0 |
| 22  | "E"   | 1630.0 |
| 23  | "E"   | 1515.0 |
| 24  | "E"   | 1635.0 |
| 25  | "E"   | 1625.0 |
| 26  | "F"   | 1520.0 |
| 27  | "F"   | 1455.0 |
| 28  | "F"   | 1450.0 |
| 29  | "F"   | 1480.0 |
| 30  | "F"   | 1445.0 |

If I wanted to read a CSV file using the facilities in R I could use

julia> rcopy("read.csv('/usr/share/distro-info/debian.csv')")
17x6 DataFrames.DataFrame
| Row | version | codename   | series | created  | release 
 | eol  |
|-|-|||--|--|--|
| 1   | 1.1 | "Buzz" | "buzz" | "1993-08-16" | 
"1996-06-17" | "1997-06-05" |
| 2   | 1.2 | "Rex"  | "rex"  | "1996-06-17" | 
"1996-12-12" | "1998-06-05" |
| 3   | 1.3 | "Bo"   | "bo"   | "1996-12-12" | 
"1997-06-05" | "1999-03-09" |
| 4   | 2.0 | "Hamm" | "hamm" | "1997-06-05" | 
"1998-07-24" | "2000-03-09" |
| 5   | 2.1 | "Slink"| "slink"| "1998-07-24" | 
"1999-03-09" | "2000-10-30" |
| 6   | 2.2 | "Potato"   | "potato"   | "1999-03-09" | 
"2000-08-15" | "2003-07-30" |
| 7   | 3.0 | "Woody"| "woody"| "2000-08-15" | 
"2002-07-19" | "2006-06-30" |
| 8   | 3.1 | "Sarge"| "sarge"| "2002-07-19" | 
"2005-06-06" | "2008-03-30" |
| 9   | 4.0 | "Etch" | "etch" | "2005-06-06" | 
"2007-04-08" | "2010-02-15" |
| 10  | 5.0 | "Lenny"| "lenny"| "2007-04-08" | 
"2009-02-14" | "2012-02-06" |
| 11  | 6.0 | "Squeeze"  | "squeeze"  | "2009-02-14" | 
"2011-02-06" | "2014-05-31" |
| 12  | 7.0 | "Wheezy"   | "wheezy"   | "2011-02-06" | 
"2013-05-04" | ""   |
| 13  | 8.0 | "Jessie"   | "jessie"   | "2013-05-04" | 
"2015-04-25" | ""   |
| 14  | 9.0 | "Stretch"  | "stretch"  | "2015-04-25" | ""   
| ""   |
| 15  | 10.0| "Buster"   | "buster"   | "2018-07-01" | ""   
| ""   |
| 16  | NA  | "Sid"  | "sid"  | "1993-08-16" | ""   
| ""   |
| 17  | NA  | "Experimental" | "experimental" | "1993-08-16" | ""   
| ""   |


(It turns out that R's allowing either ' or " for enclosing strings is an 
advantage for quoting strings within strings.)

Re: [julia-users] Re: documentation suggestions

[ivo welch]

ooops...I leaked my signature.   not a problem, but it is also was not
necessarily what I had meant to say.  for those who are interested, here is
a little background from my side of the world.

ucla anderson, like most other business schools, has been pretty ignorant
with respect to any kind of research computing expertise.

this is beginning to change, as management schools (incl us) are moving
towards one-year quantitatively oriented one-year masters program.
 anderson already has a masters of financial engineering and is about to
start a masters program in data analytics.  as for me, I am also trying to
figure out how to offer more of this to MBA students, our traditional bread
and butter, but it is not clear whether this can be implemented.  so, in
the future, we will need more data, programming, and other computing
support than we did in the past.  like every other industry.

it is exceedingly difficult to hire good programmers in a context like
our's.  universities do not pay much, for institutional reasons.
 individuals that are very good at this tend to be lured away to industry
if they are good, and non-terminable if they are bad.  a year goes by very
fast---we may find someone for one year, but then not the next.  any
program has to be prepared to run for decades.  we cannot shut down a
masters program for lack of a critical person.

our current IT department (both UCLA and Anderson) mostly handle basics,
such as the network and Microsoft apps.  as far as I can tell,
http://www.ats.ucla.edu/stat/ offers some R expertise, but not Julia
expertise.  its depth has varied with the individuals working there.  there
is no julia support afaik.

our best choices are typically individuals that want to get a phd and just
happen to have good expertise.  R, julia, etc.  another choice would be
someone who wants to work half-time on a project like julia and the other
half-time work on direct program support.  job has nice benefits...

just to get a position approved can take UC about 3-6 months and is a
high-effort affair.  we have rules up the wazoo.  there is also one month
of data expertise that anyone would want to learn (WRDS, CRSP, Compustat).
I can spend a month full-time to get there.  sigh.

so, for the most part, the few of us faculty and phd students, who like
programming have been bootstrapping it ourselves.  at UCLA Anderson, we are
luckier in this respect than many other places (Keith Chen, Peter Rossi,
John Mamer, ...), but it's tough.

julia expertise would be great for us to have.  it would have great
externalities for us.  if anyone with deep julia expertise wants to apply
to UCLA for a few years (phd, undergrad, master), with a side job at
Anderson, then drop me an email ;-).  for obvious reasons, faculty has and
wants no power to make admission decisions (or we would be besieged by our
friends and family), but I could put in a good word with our admissions
department(s).  it matters on the margin.  if someone working on julia
wants a regular job, also please email me.

/iaw

Ivo Welch (ivo.we...@gmail.com)
http://www.ivo-welch.info/


On Wed, Feb 10, 2016 at 2:47 PM, Jeffrey Sarnoff 
wrote:

> That is a reasonable want; it may take Anderson some time to institute
> scholarships for expertise in Julia
> If you were already expert with Julia, what would you have your students
> doing?
>
>
>   for expertThat is a reasonable want.  As an alternative, Anderson is
> not offering scholarships earmarked for Julia experts.
>
> On Wednesday, February 10, 2016 at 3:49:47 PM UTC-5, ivo welch wrote:
>>
>>
>> indeed.  thank you, josh.  I would add a final chapter at
>>
>> http://docs.julialang.org/en/release-0.4/
>>
>> with a set of links to various further resources, examples, full
>> stand-alone programs, etc.  for me, at least, the perl cookbook and sets of
>> self-contained snippet programs to start with, were the main reason why I
>> learned perl many years ago.
>>
>> the key problem to my use of julia over R for my students is that I do
>> not have a resident julia expert at UCLA.  this won't change anytime soon,
>> because they are hard to find (hire) :-(.  this google forum is great, but
>> it's scary to switch without a double hull.  many, many full *working*
>> standalone examples are the next best thing for me.
>>
>> regards,
>>
>> /iaw
>>
>>
>> 
>> Ivo Welch (ivo@gmail.com)
>> http://www.ivo-welch.info/
>> J. Fred Weston Distinguished Professor of Finance
>> Anderson School at UCLA, C519
>> Free Finance Textbook, http://book.ivo-welch.info/
>> Exec Editor, Critical Finance Review,
>> http://www.critical-finance-review.org/
>> Editor and Publisher, FAMe, http://www.fame-jagazine.com/
>>
>> On Wed, Feb 10, 2016 at 10:25 AM, Josh Day  wrote:
>>
>>> I think a lot of what you're looking for already exists.  It's just that
>>> things like "run a regression according to variable names" wouldn't belong
>>> in base Julia.  If you haven't already, I'd take a look at StatsBase.jl,
>>> DataF

Re: [julia-users] Re: documentation suggestions

[Jeffrey Sarnoff]

If you want to use it,  the julia-jobs forum 
 exists to let people 
know of opportunities that are posted there.

On Wednesday, February 10, 2016 at 5:47:46 PM UTC-5, Jeffrey Sarnoff wrote:
>
> That is a reasonable want; it may take Anderson some time to institute 
> scholarships for expertise in Julia
> If you were already expert with Julia, what would you have your students 
> doing?
>
>
>   for expertThat is a reasonable want.  As an alternative, Anderson is 
> not offering scholarships earmarked for Julia experts. 
>
> On Wednesday, February 10, 2016 at 3:49:47 PM UTC-5, ivo welch wrote:
>>
>>
>> indeed.  thank you, josh.  I would add a final chapter at 
>>
>> http://docs.julialang.org/en/release-0.4/
>>
>> with a set of links to various further resources, examples, full 
>> stand-alone programs, etc.  for me, at least, the perl cookbook and sets of 
>> self-contained snippet programs to start with, were the main reason why I 
>> learned perl many years ago.
>>
>> the key problem to my use of julia over R for my students is that I do 
>> not have a resident julia expert at UCLA.  this won't change anytime soon, 
>> because they are hard to find (hire) :-(.  this google forum is great, but 
>> it's scary to switch without a double hull.  many, many full *working* 
>> standalone examples are the next best thing for me.
>>
>> regards,
>>
>> /iaw
>>
>>
>> 
>> Ivo Welch (ivo@gmail.com)
>> http://www.ivo-welch.info/
>> J. Fred Weston Distinguished Professor of Finance
>> Anderson School at UCLA, C519
>> Free Finance Textbook, http://book.ivo-welch.info/
>> Exec Editor, Critical Finance Review, 
>> http://www.critical-finance-review.org/
>> Editor and Publisher, FAMe, http://www.fame-jagazine.com/
>>
>> On Wed, Feb 10, 2016 at 10:25 AM, Josh Day  wrote:
>>
>>> I think a lot of what you're looking for already exists.  It's just that 
>>> things like "run a regression according to variable names" wouldn't belong 
>>> in base Julia.  If you haven't already, I'd take a look at StatsBase.jl, 
>>> DataFrames.jl, and GLM.jl.
>>>
>>>
>>> http://dataframesjl.readthedocs.org/en/latest/io.html#importing-data-from-tabular-data-files
>>> https://github.com/JuliaStats/GLM.jl
>>>
>>>
>>>
>>> On Wednesday, February 10, 2016 at 10:58:37 AM UTC-5, ivo welch wrote:


 ladies and gents---I am not (yet) a julia user.

 may I suggest adding more examples into two places where julia users 
 will face starting hurdles?

 [1] the I/O docs of julia.  like, reading and writing csv files that 
 are compressed and decompressed on-the-fly, even if not in the ultimate 
 efficient manner.a large fraction of the time and frustration of new 
 users is consumed by the task of shoehorning data into and out of new 
 computer languages.  with all of R's problem, the ' d <- 
 read.csv("f.csv")' 
 and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry 
 frustration greatly.  perhaps xml file reading and writing.  perhaps...

 [2] more 'standard task' programs would be great.  read a csv file, run 
 a regression according to variable names on the command line, print 
 output, 
 draw a graph.  I know there are fragments throughout the docs, but some 
 section with ready to run complete programs would be good, perhaps at the 
 end of the manual.

 in a year, I hope to switch my students from R to julia.

 regards,

 /iaw


>>

Re: [julia-users] Re: documentation suggestions

[Jeffrey Sarnoff]

That is a reasonable want; it may take Anderson some time to institute 
scholarships for expertise in Julia
If you were already expert with Julia, what would you have your students 
doing?


  for expertThat is a reasonable want.  As an alternative, Anderson is not 
offering scholarships earmarked for Julia experts. 

On Wednesday, February 10, 2016 at 3:49:47 PM UTC-5, ivo welch wrote:
>
>
> indeed.  thank you, josh.  I would add a final chapter at 
>
> http://docs.julialang.org/en/release-0.4/
>
> with a set of links to various further resources, examples, full 
> stand-alone programs, etc.  for me, at least, the perl cookbook and sets of 
> self-contained snippet programs to start with, were the main reason why I 
> learned perl many years ago.
>
> the key problem to my use of julia over R for my students is that I do not 
> have a resident julia expert at UCLA.  this won't change anytime soon, 
> because they are hard to find (hire) :-(.  this google forum is great, but 
> it's scary to switch without a double hull.  many, many full *working* 
> standalone examples are the next best thing for me.
>
> regards,
>
> /iaw
>
>
> 
> Ivo Welch (ivo@gmail.com )
> http://www.ivo-welch.info/
> J. Fred Weston Distinguished Professor of Finance
> Anderson School at UCLA, C519
> Free Finance Textbook, http://book.ivo-welch.info/
> Exec Editor, Critical Finance Review, 
> http://www.critical-finance-review.org/
> Editor and Publisher, FAMe, http://www.fame-jagazine.com/
>
> On Wed, Feb 10, 2016 at 10:25 AM, Josh Day  > wrote:
>
>> I think a lot of what you're looking for already exists.  It's just that 
>> things like "run a regression according to variable names" wouldn't belong 
>> in base Julia.  If you haven't already, I'd take a look at StatsBase.jl, 
>> DataFrames.jl, and GLM.jl.
>>
>>
>> http://dataframesjl.readthedocs.org/en/latest/io.html#importing-data-from-tabular-data-files
>> https://github.com/JuliaStats/GLM.jl
>>
>>
>>
>> On Wednesday, February 10, 2016 at 10:58:37 AM UTC-5, ivo welch wrote:
>>>
>>>
>>> ladies and gents---I am not (yet) a julia user.
>>>
>>> may I suggest adding more examples into two places where julia users 
>>> will face starting hurdles?
>>>
>>> [1] the I/O docs of julia.  like, reading and writing csv files that are 
>>> compressed and decompressed on-the-fly, even if not in the ultimate 
>>> efficient manner.a large fraction of the time and frustration of new 
>>> users is consumed by the task of shoehorning data into and out of new 
>>> computer languages.  with all of R's problem, the ' d <- read.csv("f.csv")' 
>>> and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry 
>>> frustration greatly.  perhaps xml file reading and writing.  perhaps...
>>>
>>> [2] more 'standard task' programs would be great.  read a csv file, run 
>>> a regression according to variable names on the command line, print output, 
>>> draw a graph.  I know there are fragments throughout the docs, but some 
>>> section with ready to run complete programs would be good, perhaps at the 
>>> end of the manual.
>>>
>>> in a year, I hope to switch my students from R to julia.
>>>
>>> regards,
>>>
>>> /iaw
>>>
>>>
>

Re: [julia-users] Re: documentation suggestions

[ivo welch]

indeed.  thank you, josh.  I would add a final chapter at

http://docs.julialang.org/en/release-0.4/

with a set of links to various further resources, examples, full
stand-alone programs, etc.  for me, at least, the perl cookbook and sets of
self-contained snippet programs to start with, were the main reason why I
learned perl many years ago.

the key problem to my use of julia over R for my students is that I do not
have a resident julia expert at UCLA.  this won't change anytime soon,
because they are hard to find (hire) :-(.  this google forum is great, but
it's scary to switch without a double hull.  many, many full *working*
standalone examples are the next best thing for me.

regards,

/iaw



Ivo Welch (ivo.we...@gmail.com)
http://www.ivo-welch.info/
J. Fred Weston Distinguished Professor of Finance
Anderson School at UCLA, C519
Free Finance Textbook, http://book.ivo-welch.info/
Exec Editor, Critical Finance Review,
http://www.critical-finance-review.org/
Editor and Publisher, FAMe, http://www.fame-jagazine.com/

On Wed, Feb 10, 2016 at 10:25 AM, Josh Day  wrote:

> I think a lot of what you're looking for already exists.  It's just that
> things like "run a regression according to variable names" wouldn't belong
> in base Julia.  If you haven't already, I'd take a look at StatsBase.jl,
> DataFrames.jl, and GLM.jl.
>
>
> http://dataframesjl.readthedocs.org/en/latest/io.html#importing-data-from-tabular-data-files
> https://github.com/JuliaStats/GLM.jl
>
>
>
> On Wednesday, February 10, 2016 at 10:58:37 AM UTC-5, ivo welch wrote:
>>
>>
>> ladies and gents---I am not (yet) a julia user.
>>
>> may I suggest adding more examples into two places where julia users will
>> face starting hurdles?
>>
>> [1] the I/O docs of julia.  like, reading and writing csv files that are
>> compressed and decompressed on-the-fly, even if not in the ultimate
>> efficient manner.a large fraction of the time and frustration of new
>> users is consumed by the task of shoehorning data into and out of new
>> computer languages.  with all of R's problem, the ' d <- read.csv("f.csv")'
>> and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry
>> frustration greatly.  perhaps xml file reading and writing.  perhaps...
>>
>> [2] more 'standard task' programs would be great.  read a csv file, run a
>> regression according to variable names on the command line, print output,
>> draw a graph.  I know there are fragments throughout the docs, but some
>> section with ready to run complete programs would be good, perhaps at the
>> end of the manual.
>>
>> in a year, I hope to switch my students from R to julia.
>>
>> regards,
>>
>> /iaw
>>
>>

[julia-users] Re: documentation suggestions

[Josh Day]

I think a lot of what you're looking for already exists.  It's just that 
things like "run a regression according to variable names" wouldn't belong 
in base Julia.  If you haven't already, I'd take a look at StatsBase.jl, 
DataFrames.jl, and GLM.jl.

http://dataframesjl.readthedocs.org/en/latest/io.html#importing-data-from-tabular-data-files
https://github.com/JuliaStats/GLM.jl



On Wednesday, February 10, 2016 at 10:58:37 AM UTC-5, ivo welch wrote:
>
>
> ladies and gents---I am not (yet) a julia user.
>
> may I suggest adding more examples into two places where julia users will 
> face starting hurdles?
>
> [1] the I/O docs of julia.  like, reading and writing csv files that are 
> compressed and decompressed on-the-fly, even if not in the ultimate 
> efficient manner.a large fraction of the time and frustration of new 
> users is consumed by the task of shoehorning data into and out of new 
> computer languages.  with all of R's problem, the ' d <- read.csv("f.csv")' 
> and 'd<-read.csv(pipe(paste("gzcat ", fname)))' reduced this entry 
> frustration greatly.  perhaps xml file reading and writing.  perhaps...
>
> [2] more 'standard task' programs would be great.  read a csv file, run a 
> regression according to variable names on the command line, print output, 
> draw a graph.  I know there are fragments throughout the docs, but some 
> section with ready to run complete programs would be good, perhaps at the 
> end of the manual.
>
> in a year, I hope to switch my students from R to julia.
>
> regards,
>
> /iaw
>
>