[R-SIG-Finance] Reading MetaStock data format in R

[Paul Gilbert]

Peter

(No doubt you are working in a context where you don't have control or 
influence over everything, but ... )

I think you are really attacking the wrong part of your problem. If you 
are going to all the trouble of downloading and updating daily data on 
an inhouse db, why don't you put it in a modern open format that is 
generally usable, or even put it in a real database?

Best,
Paul


On 13-09-30 10:19 AM, R. Michael Weylandt <michael.weylandt at gmail.com>
wrote:
>
>
> On Sep 30, 2013, at 9:55, Peter Fertig <pfertig at yahoo.de> wrote:
>
>> Hi Michael,
>>
>> thank you for your comments. I have learnt now that using the DLL
>> file, will not work as I can not modify the DLL.
>
> You don't need to modify it -- just wrap it in something which
> 'speaks R' rather than trying to call it directly.
>
> M
>
>
>>
>> Creating csv or Excel files, another proposal I got, is what I have
>> to do already for another commercial econometric software (RATS
>> from Estima) I use now for a couple of years. Thus, I limited this
>> to weekly data for only a few instruments. However, I would like to
>> run the analysis on a daily basis for more instruments. Therefore,
>> I still look for a solution in R by using C-code. Unfortunately, I
>> have only experience in programming in Visual Basic and
>> Pascal/Delphi. Maybe having only the code for the functions is not
>> sufficient and the C-file needs a few more lines? Something, I
>> still have to explore.
>>
>> Best regards Peter
>>
>> Von: "R. Michael Weylandt <michael.weylandt at gmail.com>"
>> <michael.weylandt at gmail.com> An: Peter Fertig <pfertig at yahoo.de>
>> CC: "r-sig-finance at r-project.org" <r-sig-finance at r-project.org>
>> Gesendet: 14:45 Montag, 30.September 2013 Betreff: Re:
>> [R-SIG-Finance] Reading MetaStock data format in R
>>
>>
>>
>> On Sep 30, 2013, at 3:28, Peter Fertig <pfertig at yahoo.de> wrote:
>>
>>> Hello,
>>>
>>> I am new at R and need some help because I am stuck in a project.
>>> I have financial data, which is stored in the MetaStock data
>>> format and is updated daily. For reading this data, I started to
>>> write a R-script. So far, I managed to read the Master file,
>>> which contains the information about the various securities in a
>>> directory, as well as reading the Fxxx.Dat data files. The data
>>> is stored in the Microsoft Binary Format, which was used for
>>> Basica under DOS. Thus, it has to be converted into the IEEE
>>> format for the floating point numbers. But here my problems
>>> start.
>>
>> Godspeed -- nothing quite like non-standard binary formats before
>> the first morning coffee.
>>
>>
>>>
>>> I have a DLL-file with functions to convert the 4byte MBF format
>>> into IEEE format as well as from IEEE to MBF format. As
>>> recommended, this DLL is in the Windows/System32 directory. I use
>>> this DLL with Visual Basic and it works fine. In the R script, I
>>> included the following line
>>>
>>> dyn.load("MBFIEE32", DLLpath="C:/Windows/System32").
>>
>> Don't think it works like that -- anything you want exposed within
>> R will need to return R objects or wrapper classes (SEXP's
>> generally) -- if your DLL works for / was written for VB, it won't
>> work un-modified in R.
>>
>>>
>>> However, testing if the DLL had been loaded with
>>> is.loaded("MBFIEE32") returns FALSE. I have read the manual, but
>>> have not figured out what went wrong. Thus, my question is, what
>>> might I have overlooked, which could cause this error?
>>>
>>> I also have a file with the functions for the format conversions
>>> in C. After making slight modifications to ensure that the
>>> functions return void as requested, I wanted to compile the file
>>> with the command (using the shell feature of RStudio) R CMD SHLIB
>>> [-d] -o MBF2IEEE.DLL MBF2IEEE.c. I called this command from the
>>> directory where the C-file is stored. However, the file was not
>>> compiled. Does the C-file has to be stored in a particular
>>> directory? After reading the online help, I can not figure out,
>>> what went wrong.
>>>
>>
>> Is your R development environment set up properly? R is not
>> officially supported to build under VS and I'd imagine the same
>> applies to packages.
>>
>> To test that, I'd try building the 'xts' package locally -- it's an
>> easy (enough) build and relevant for finance work anyways.
>>
>> Depending on the scale of your data and the frequency you'll be
>> doing this, it might be worth writing a standalone program which
>> makes a csv file (or similar) and then just reading that into R so
>> you can get on with your work.
>>
>> Cheers, Michael
>>
>>
>>> Thanks for any help
>>>
>>> Peter [[alternative HTML version deleted]]
>>>
>>> _______________________________________________
>>> R-SIG-Finance at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-finance --
>>> Subscriber-posting only. If you want to post, subscribe first. --
>>> Also note that this is not the r-help list where general R
>>> questions should go.
>>
>>
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> R-SIG-Finance at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-finance --
> Subscriber-posting only. If you want to post, subscribe first. --
> Also note that this is not the r-help list where general R questions
> should go.
>