SteadyOptions is an options trading forum where you can find solutions from top options traders. TRY IT FREE!

We’ve all been there… researching options strategies and unable to find the answers we’re looking for. SteadyOptions has your solution.

gf58

GF58's Intraday RV monitor for Interactive Brokers

213 posts in this topic

Recommended Posts

20 hours ago, J_D said:

I cannot seem to downgrade matplotlib either to 3.2.2 or 3.1.3, neither from Anacondas GUI not command prompt. It waits and waits and finally errors out.

From what I understand there's a couple of different options for package managers. conda/anaconda is the better to try in the first instance for reasons that I can no longer remember. However, sometimes it gets the list of dependencies tangled up (this requires that which requires that other thing that needs only this version of....etc)....which is what that looks like.

Im sure that this means something important but as a man of action, when that happens I just take a different hammer out of the tool box and resume hammering:

pip install matplotlib=3.2.2

 

Try that from the anaconda prompt. pip is another package manager and often works for me when the conda install hangs on me....apparently this isnt a good idea to have two different tools managing dependencies but hey, its already tangled up by the time you're turning to pip.

Share this post


Link to post
Share on other sites
1 hour ago, gf58 said:

From what I understand there's a couple of different options for package managers. conda/anaconda is the better to try in the first instance for reasons that I can no longer remember. However, sometimes it gets the list of dependencies tangled up (this requires that which requires that other thing that needs only this version of....etc)....which is what that looks like.

Im sure that this means something important but as a man of action, when that happens I just take a different hammer out of the tool box and resume hammering:

pip install matplotlib=3.2.2

 

Try that from the anaconda prompt. pip is another package manager and often works for me when the conda install hangs on me....apparently this isnt a good idea to have two different tools managing dependencies but hey, its already tangled up by the time you're turning to pip.

That worked senor @gf58 Charts are fixed too. Many thanks.

Share this post


Link to post
Share on other sites
20 hours ago, gf58 said:

 I just take a different hammer out of the tool box

Not different hammer, BIGGER hammer.  Got to get your priorities straight.  Think "Terminator" with the conflicting programming smashing a Ford Explorer (or whatever it was) into paste.

Share this post


Link to post
Share on other sites
55 minutes ago, Ringandpinion said:

Not different hammer, BIGGER hammer.  Got to get your priorities straight.  Think "Terminator" with the conflicting programming smashing a Ford Explorer (or whatever it was) into paste.

Good point. There's no glory in using WD40 on a stuck joint or figuring out why the package wasnt working. Otherwise we'd be reading those manuals in Ikea boxes before we get hopelessly stuck (as opposed to quietly in a discrete corner out of the view of others in the room).

Share this post


Link to post
Share on other sites

FYI, several weeks ago I had installed the standalone TWS API to try out the chartaffair intraday tool.  It didn't work so I didn't fool with it any further.  Friday the gf58ometer quit working.  Uninstalled the standalone TWS API and reinstalled TWS Latest just to be sure.  All is well.

Share this post


Link to post
Share on other sites
On 2/20/2021 at 4:30 AM, Ringandpinion said:

So far today, I have not gotten a single functional graph, always missing many chunks.  Anybody else experiencing this behavior?

Time of day seems to have a huge impact.  I've always found that retrieving whole histories of charts during the session to be problematic. I generally refresh all my charts about 2 hours prior to open and then do a couple of updates around mid session. As an example, if I try to refresh 20 charts on my watchlist 2 hours prior to open I can get that done in about 20-30 minutes. If I try to do that during session it used to take the better part of the morning. I also find calendars will never return new chunks in the first 30-60 minutes of the session.

 

That said, I did notice some issues late last week (I was trying to get that AVGO calendar for the thread but kept receiving junk data).

Share this post


Link to post
Share on other sites

New release 1.2.07 with a minor fix for calendar errors when new strikes get added.

  • Due to my putting a < around the wrong way caused it to ask for calendars on strikes that didnt exist.
  • This only popped up for the couple of hours on the day after new strikes have been announced but are yet to be traded
  • Of course, fixing it wasn't just a matter of turning the symbol around the wrong way, it also needed an easy way to flush the bad data out of database

 

So, if you're seeing errors like (bid cant be transmitted due to missing leg):

image.png

when loading calendars, after updating the files and restarting, call:

rvf.flush('XXX') where XXX is the chart name. You'll get a rather generic 'Done' message when its completed flushing the bad data from the database. Now you should be able to rvf.load('XXX') as normal without the missing chunks from calendars that dont exist.

 

 

On the other hand, everyone should update to this version when they get a chance just to prevent the issue occurring in the first place.

Note that this time you'll need to overwrite both the RVFactory.py and LocalConnection.py files 

 

Share this post


Link to post
Share on other sites
8 minutes ago, Ringandpinion said:

It is likely I'm the only one using your colorblind pallete but just for the record, the matplotlib I downgraded to doesn't work with it.  I upgraded to (version 3.2.2) just before the busted package and it worked again.

Id guess its probably something to do with slightly different predefined colour names. So its all working at the moment (complete with minor labels)?

 

NB. Funny that you mention it today. I was teaching a class this afternoon and started filling in specific details on a diagram using different coloured whiteboard markers (colour coding them) to make it easier to follow...and then I looked at the audience, noticed the entirely male faces staring back at me, estimated that statistically one of them must be and then stopped myself to check with the group if anyone was colour blind. 

In other words, even if you've got the feature to yourself its certaintly made me more aware of of how it works/how to design around it at least.

....but very coincidental you mentioned that today!

Share this post


Link to post
Share on other sites

It is an odd condition.  It has some disadvantages but some definite advantages.  Since I've lived my whole life colorblind, it is normal.  There are some forms of colorblindness that are serious disabilities, mine is just the run of the mill type.  I don't see reds as well as others.  But the brain compensates and I see things others don't because I distinguish shade and pattern differently.  Buy me a beer one day and I'll tell you about it, it is really kind of interesting.  The organic machine we call a body is fascinating in it's function and complexity.  And I don't know of a self repairing machine that man has made yet and this one lasts the better part of a century unless you were born a century ago.

Share this post


Link to post
Share on other sites
21 hours ago, gf58 said:

Time of day seems to have a huge impact.  I've always found that retrieving whole histories of charts during the session to be problematic.

I wondered if the data was more sporadic during the regular session.  It seemed so but without testing it definitively I just thought it gave more errors.  Your confirmation is appreciated.  Unfortunately, I am not in the habit of identifying trade candidates early enough to take advantage of the better API flow after hours, something I may need to change.  The gf58ometer really filled a need that I had a trouble finding a solution to.  After all the analysis, I really wanted something to help with trade entry targeting so I didn't get in on a high point.  I have put a lot of effort into it and then you solved the problem.  Thanks again.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.