Can't Locate Community.Components
Author: mosherif
Creation Date: 12/5/2009 11:56 AM
profile picture

mosherif

#1
Apologies for the previous misplaced post. While trying to run the sample code for FillSeriesFromFile:

CODE:
Please log in to see this code.


Getting the following error message:

error CS0234@(7, 17): Type or namespace 'Component' does not exist in the namespace 'Community'.

What might be causing this?

Thanks.
profile picture

Eugene

#2
Hi,

You're missing the Community.Components library.

Hmm, I thought it's explained pretty clear on the library's home page [in our Wiki, where you found FillSeriesFromFile]:
QUOTE:
The Extension file is available to download from our site (Wealth-Lab.com) after registration and logging in.


Click on "Extensions" section, review the tutorial, then "Get Extensions", and find Community.Components there.
profile picture

mosherif

#3
Thanks, everything is now working, however the series plots as a zero line, although the scale on the left seems appropriate, running from 0 to 1.00. Here is a sample of the first few lines of the text file. Any ideas why?

20081117,0.7071
20081118,0.4064
20081119,0.1317
20081120,0.0609
20081121,0.2361
20081124,0.6558
20081125,0.8999
20081126,0.9711
20081128,0.5188
20081201,-0.5907
20081202,-0.9928
20081203,-0.8154
profile picture

Eugene

#4
I see a histogram of your values plotted perfectly. What's the data range of your active strategy window?
profile picture

mosherif

#5
Thanks Eugene,

If I start with a folder that was originally setup for inter day data and then change the periodicity to daily to match the data in the text file, the external data window does not seem to refresh. However, if you start with a security that was originally created with daily data, everything works perfectly.

Thanks for your help.
profile picture

Eugene

#6
That's daily series data in your example above. The method is not specifically tweaked for synchronizing series in different time frames, so it assumes you'll be using the same scale (here: daily).

I think you could have seen the values (just one per day) if your intraday data were a futures trading round the clock. For example, "20081117" is actually treated as "20081117 0:00" so you'd probably saw "0.7071" for the 12 AM bar in this case.
profile picture

mosherif

#7
Also trying various combination of sample data for 30 min bars, but getting a run-time error. What am I missing?

Here is one sample:

CODE:
Please log in to see this code.



20080929,-0.71428 10:00 am
20080929,0.5000 10:30 am
20080929,-0.4615 11:00 am
20080929,0.8769 11:30 am

Thanks.
profile picture

Eugene

#8
"yyyyMMdd hh:mm tt" is not equal to "20080929,-0.71428 10:00 am" and won't be parsed. "20080929 10:00 am,series value" probably is.
profile picture

mosherif

#9
Sorry for flipping that around and putting the date at the end. I tried a few more alternatives, but still no luck. Here is one that is still giving a

Runtime error: Input string was not in a correct format.

CODE:
Please log in to see this code.


20080929 10:00,-0.71428
20080929 10:30,0.5000
20080929 11:00,-0.4615
20080929 11:30,0.8769

Thanks for your help.
profile picture

Eugene

#10
Sorry, you're right -- something's wrong with the overloaded call and getting intraday series to work. I'll take a look.
profile picture

mosherif

#11
Thanks for all your help on this.
profile picture

Eugene

#12
Thank you for the heads-up.

I found a bug in the overloaded function that was preventing it from parsing anything more complex than "yyyyMMdd" or alike.

A fix will be available shortly, in release 2009.12 of Community.Components.
profile picture

mosherif

#13
Thanks, will be looking out for it.
profile picture

Eugene

#14
Community.Components 2009.12 was released today; please try.
profile picture

mosherif

#15
Everything is working as expected and the refresh issue when switching periodicity has also been addressed. Thanks for your help and the incredibly fast turnaround!
This website uses cookies to improve your experience. We'll assume you're ok with that, but you can opt-out if you wish (Read more).