Closed grumpyinca closed 2 years ago
The tutor7 script calls seek twice causing it to overlay the first seek's message with the second seek's message.
title: "Page 06 of 17",
text: (
<>
<p>
If you see the Feasibility status shown as "NOT FEASIBLE",
It means that the solution point has slightly violated constraints
in order to achieve improvements in the result
(minimum Weight in this case).
</p>
<p>
You may see the message:
<br />
"TO FURTHER IMPROVE RESULT, RE-EXECUTE SEEK"
<br />
This can happen when the optimum is far from the starting point.
As a result, the internal estimate of the optimum was not perfect.
Simply re-execute Seek to refine the result.
</p>
<p>
Let's try exploring in a different direction.
Perhaps this will be a design that prioritizes a softer spring
with a reduced spring rate.
Take note of the current value of Rate.
</p>
<p>
In the process of moving to the next page
the tutorial will execute:
<br /><br />
SEEK MIN Rate
</p>
<br />
</>
),
actions: [
seek("Weight", MIN),
seek("Weight", MIN)
]
Re-running with manually issuing first one seek, then a search, then another seek does the following
This is not a programming issue, instead is it a tutor7 script issue. Closing and marking as invalid
The Seek algorithm conducts a second search in the case where the first search does not produce a feasible result.
By observation of the termination message and console log output (see below) it is apparent that current coding resets the reference point used in the calculation of % improvement with each search. Thus, in those cases where the Seek algorithm requires second search, the stated % improvement can be significantly incorrect.
Example using tutor7 p5: