+1 888 340 6572
MultiCharts Project Management
previous_open_issue.png
Go to the previous open issue
previous_issue.png
Go to the previous issue (open or closed)
star_faded.png
Please log in to bookmark issues
feature_request_small.png
Open Feature request MC-162

Need to Allow Step value of 0 (Support for Sequential Optimization)

action_vote_minus_faded.png
3
Votes
action_vote_plus_faded.png
next_issue.png
Go to the next issue (open or closed)
next_open_issue.png
Go to the next open issue
Description

Originally from the Forum:

http://www.multicharts.com/discussion/viewtopic.php?f=1&t=6621

Related to (but not the same as) the idea of having an on/off checkbox
for each optimization input is the idea of allowing Step = 0 as a way to
tell the optimizer "stay at the start value".
Doing sequential optimizations (first some params, then different params) with MC is currently agonizing.
IMO, this should be dead-simple to do, with every possible form of support provided for.
On/Off checkboxes and allowing Step = 0 are two of the desirable forms of support.
The problem with requiring Step 0 as currently, is that it is not possible to prevent optimization of a given parameter without destroying the **(normal, default, standard, generally used, special case, - pick one) **range settings for the parameter.
This is a big-time drag for doing sequential optimizations.
This is hopefully very simple to fix, with just minor design changes.

Steps to reproduce this issue

Current behavior does not allow step value of 0.
What is sequential optimization?
Start with opt parms set to "I believe these values should be close to good" settings
Take one of the parms and vary it across the local range (e.g. 10 to 100 iterations).  Best result may be different than the starting setting.   This becomes the new baseline.
Repeat with a different parameter.
In just a few runs, you have an excellent idea whether your strat works as intended, and what settings show favorably for the market/timeframe under test.
This is THE FASTEST WAY (and desirably interactive) to do first-cut testing.  No, you don't cover the entire search space this way.  No, you don't find the absolute peak this way.   That's OK, those are different operations.
This request is about MC's current severe disability compared to other platforms when it comes to doing sequential optimization as quick first cut testing.   IOW, it's about the user not being able to get a quick interactive answer when he should be able to!
Note:  after each run, the next run needs to be relative to the new baseline.  If this is not clear, contact me so I can clarify it for you.  Thx.

Comments (2)
#0
user-offline.png  MC_Prog (MC_Prog)
Dec 30, 2011 - 14:52

I'm still missing this in a big way ...

#0
user-offline.png  MC_Prog (MC_Prog)
Apr 13, 2012 - 16:51

The great speed of MC optimization is undercut by this workflow obstacle.  Please remedy this!

History
Issue basics
  • Type of issue
    Feature request
  • Category
    Usability
  • Targeted for
    Not determined
  • Status
    Under Review
User pain
  • Type of bug
    Not triaged
  • Likelihood
    Not triaged
  • Effect
    Not triaged
Affected by this issue (3)
People involved
Times and dates
  • Posted at
  • Last updated
Issue details
  • Resolution
    Not determined
  • Severity
    Normal
Attachments (0)
There is nothing attached to this issue
Commits (0)
There are no code checkins for this issue
Duplicate issues (0)
This issue does not have any duplicates