Back to General discussions forum
I'm trying to solve problem #169, but I'm stuck "tuning" parameters to gradient descent however sometimes I get correct parameters that works with several ships, the problem is that although in my simulation, which I probe with the debugging data that was given to us in the prblem, I get values of speed under 5 m/s when I give my responses to the verifier my ships always crash, I want to know if the verifier check the time of simulation and if that is the case what is the maximum time of simulation.
PD. If someone can give me a hint for tuning that would be good
For example for this ship (4800 7900 240000 1600)
I got this (14.859603947249536 34.451474802141895 -0.0803315740334831) my speed was 3.968397116691322 m/s
And the verifier gave me this: Ship #1 landed with speed 309.376755214 and crashed!
For example for this ship (5700 7100 350000 1560)
I got this (145.7320930791691 14.198757949820688 1.2551266529169847) my speed was 2.4062865201670642 m/s
And the checker gave me: Ship #1 crashed!
As far as I can remember, I hit similar problem. In my case I revealed, that at the supposed end of trip my ship had plenty of unused fuel. Also, my simulation algorithm was slightly changed to increase accuracy of integration. But if calculate is done with rough algorithm given on this site, ship is decelerated to zero speed too early, had not got touchdown. So simulation is not stopping and is going ahead. Ship is flying away, gaining speed, spending all remaining fuel and crashing back in free fall.
This was my scenario giving me weird big speeds of crash.
Ummmm I gonna review my accuracy on integration maybe that could be my problem, however in my case zero altitute is always reached, the unique odd thing in my case will be unused fuel.
Now, I don't now what is happening. I made that my code calculate parameters for the given example in challenge, in this case only with the first one.
Ship given as an exmaple in the problem: 6500 7600 200000 1660
I got this:
tf: 64.08520973982843
rate0: 23.80530295463453
drate: 2.218850645839510
Landing time: 146.5300000000117 s
Altitute just before landing 0.020754480965967444
Velocity of landing: 4.877396369573798
If you see, I get a lower time of landing compared with the example, then in this moment I think that this could be a problem with the checker but also a coincidence..
Ughhh man hahaha, I just found the problem...
The tf value has to be an integer. I didn't realize that rule.