[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [amibroker] Re: In the Ongoing Discussion of Historical Fundamental Data



PureBytes Links

Trading Reference Links

> How to go from there to an SQL database is the next piece of the 
> puzzle.
> That's the bit I can't do so far
> I was just about to get the SQL download and have a read.

As far as open source (free) solution is considered:

Using open source mySQL database:
http://dev.mysql.com/doc/refman/4.1/en/load-data.html

mySQL + PHP + phpMyAdmin (for easy administering of mySQL server):
http://easyphp.org/

Best regards,
Tomasz Janeczko
amibroker.com
----- Original Message ----- 
From: "brian.z123" <brian.z123@xxxxxxxxxxxx>
To: <amibroker@xxxxxxxxxxxxxxx>
Sent: Saturday, July 22, 2006 11:27 AM
Subject: [amibroker] Re: In the Ongoing Discussion of Historical Fundamental Data


> You're welcome.
> 
> Sometimes in trading it is actually an advantage not to know too 
> much, in advance, about mathematics or computers.
> If you don't you really have to make absolutely sure you understand 
> each step and brutally reduce everything down to it's most simple 
> explanation.
> It's amazing the things you find out when you do that.
> It takes a bit longer but in the end I think technoNOTS, like me, 
> can understand it as well as anybody.
> 
> I'm assuming you mean you want to take the Yahoo fdata you download 
> into AB and save it somewhere to make a histo.
> 
> I haven't walked all the way through that one as I am going a 
> slightly different route.
> Still not to say I won't want to do that for some reason at some 
> time so I am interested to find out for myself.
> 
> Here is one place we can start thinking about a possible solution to 
> the problem.
> 
> To export the fdata use explorer.
> Make a .afl explorer template for 'exporting  Yahoofdata' with
> 
> ProfitMargin = GetFnData("ProfitMargin");
> AddColumn (ProfitMargin, "ProfitMargin",1.2);
> 
> type statements for every available fdata field or every field that 
> you want (as you did in the example explorer code you posted last 
> week).
> Play around with the filter/filterNOT to get the symbol list you 
> want (all symbols?)
> Run the exploration and use the explorer export to send the fdata to 
> a .CSV file.
> Save as Don's funtastic fundas.
> The outcome is a .CSV for the day with all symbols in one file.
> The symbols are in rows and the column headings = "ProfitMargin" etc.
> If you repeat, say each day etc, you will end up with a series of 
> daily files e.g July222006, July232006 containing all symbols 
> (obviously).
> 
> That is where things start to get a little more interesting.
> How to go from there to an SQL database is the next piece of the 
> puzzle.
> That's the bit I can't do so far
> I was just about to get the SQL download and have a read.
> 
> If you get stuck from there come back with another question.
> I might be hanging around, have another go and get lucky or some 
> kind angel may be hovering above us in Ami heaven.
> 
> If you crack the D'AMI CODE let me know how to do it.
> 
> BrianB.
> 
> 
> --- In amibroker@xxxxxxxxxxxxxxx, "Don Lindberg" <dlindber@xxx> 
> wrote:
>>
>>  Brian,
>>  
>> Thanks for the reply. I now have a much clearer picture of how the 
> data is
>> kept in AB. Now I have to figure out how to create a SQL database 
> with the 
>> extra" data I want to build history on  |;>).
>>  
>> Thanks Again,
>>  
>> Don
>>  
>> -------Original Message-------
>>  
>> From: brian.z123
>> Date: 07/21/06 16:08:47
>> To: amibroker@xxxxxxxxxxxxxxx
>> Subject: [amibroker] Re: In the Ongoing Discussion of Historical 
> Fundamental
>> Data
>>  
>> Hello Don,
>>  
>> I still have a lot to learn about AB, so, with some qualification 
> the
>> answer is no.
>>  
>> Ami does not work like that.
>> It can only keep six fields or columns of data (OHLCVOI), if we
>> compare it a spreadsheet or a database table for learning purposes.
>> The other information like the company name, category and the new
>> fundamental data etc is only stored once.
>> If you are used to spreadsheets just consider that the Yahoo data 
> is
>> stored in columns with only one row, while the OHLCVOI data can be
>> stored in an unlimited number of rows (technically speaking this in
>> not correct as Ami uses arrays not 'columns' but that is how I 
> think
>> about it for ease of learning).
>> When you update your fdata with a new download from Yahoo the old
>> numbers will be replaced (overwritten) by the new numbers so there 
> is
>> No way to refer back to previous, say EPS.
>>  
>> There was a way around this by importing with the ASCII wizard and
>> using an artificial ticker to store the historical data, but that 
> is a
>> little more difficult to do.
>> To manage that with the new Yahoo fdata you would first need to 
> export
>> the data from AB on a regular basis and save each line of export by
>> appending it to the previous line(s)in a separate file outside of 
> AB
>> e.g. a .CSV file.
>> From there you can import it back into AB as an artificial ticker.
>> That is a separate subject if you want to go into it.
>>  
>> That is not so relevant as of today as Tomasz just posted a notice 
> to
>> say that the new 'database' plug in has been released.
>> It is designed to allow users to read databases containing data 
> that
>> is extra to the OHLCVOI fields that comprise the normal AB 
> databases.
>> I haven't downloaded and read the instructions as yet but I expect 
> it
>> to be easy to use.
>>  
>> In the long run, that would be the recommended way to go for
>> fundamentalists.
>> It will definitely be much easier to use historical fundamental 
> data.
>> That leaves us with the problem now of finding some databases to
>> practise on.
>> For actual trading use databases that have been kept for us by 
> others
>> are the best bet but some users will export from AB/Yhaoo data and
>> compile historcal records for themselves.
>>  
>> For commercial sources refer to mmy previous historical fundamental
>> data provider lists.
>>  
>> One of the sources in that list is reported to have free data 
> (HOOVER)
>> which would be good for training use or people who prefer that.
>> I will look more closely at the prospects on the list later.
>> I am not optimistic about the chance of finding free histo fdata 
> (the
>> cost of compiling and maintaining such databases is too high for
>> owners to give a lot of it away).
>>  
>> Some Amifriends might have histo database files they will share for
>> training use.
>>  
>> I am sure there will be more forum help forthcoming on the new 
> plugin
>> and how to acquire and use histo fundamental databases.
>>  
>> BrianB.
>>  
>> --- In amibroker@xxxxxxxxxxxxxxx, "Don Lindberg" <dlindber@> wrote:
>> >
>> > I have a question about setup of AmiBroker regarding the 
> Fundamental
>> > Data we can now dowload from Yahoo.  While I realize that Yahoo 
> does
>> > not keep this info in a Historical database, does AmiBroker?  
> What I
>> > mean is this. If I download the fundamental data, say once a 
> week,
>> am
>> > I overtime building a history of this data? And if I am, how do I
>> > compare today with last week?
>> >
>> > Don Lindberg
>> >
>>  
>>  
>>  
>>  
>>  
>>  
>> ------------------------ Yahoo! Groups Sponsor --------------------
> ~-->
>> Great things are happening at Yahoo! Groups.  See the new email 
> design.
>> http://us.click.yahoo.com/TISQkA/hOaOAA/yQLSAA/GHeqlB/TM
>> -------------------------------------------------------------------
> -~->
>>  
>> Please note that this group is for discussion between users only.
>>  
>> To get support from AmiBroker please send an e-mail directly to
>> SUPPORT {at} amibroker.com
>>  
>> For other support material please check also:
>> http://www.amibroker.com/support.html
>>  
>>  
>> Yahoo! Groups Links
>>
> 
> 
> 
> 
> 
> 
> Please note that this group is for discussion between users only.
> 
> To get support from AmiBroker please send an e-mail directly to 
> SUPPORT {at} amibroker.com
> 
> For other support material please check also:
> http://www.amibroker.com/support.html
> 
> 
> Yahoo! Groups Links
> 
> 
> 
> 
> 
> 
> 
> 
>