Graphs - Times to Deadlock

[geraintpalmer]

For the 2 node network, simulating for 1000 iterations: Base parameters: L1 = 4.0 L2 = 5.0 mu1 = 10.0 mu2 = 8.0 r12 = 0.25 r21 = 0.15 n1 = 3 n2 = 2

• [x] L1, 2.5-14 in steps of 0.5
• [x] L2, 2.5-14 in steps of 0.5
• [x] mu1, 2.5-14 in steps of 0.5
• [x] mu2, 2.5-14 in steps of 0.5
• [x] r12, 0.1-1.0 in steps of 0.05
• [x] r21, 0.1-1.0 in steps of 0.05
• [x] n1, 0-6 in steps of 1
• [x] n2, 0-6 in steps of 1

[geraintpalmer]

Varying L1

[geraintpalmer]

Varying r12

[drvinceknight]

Looking glorious.

MOAR!

[geraintpalmer]

Varying mu1

[drvinceknight]

Not as pretty... Run for longer still? How long are these going for?

[geraintpalmer]

1000 iterations.

[drvinceknight]

Would it cost much time to multiply that by 10?

[geraintpalmer]

That one with 1000 itrs had been running since 9:30, just finished.

[drvinceknight]

Ok, get a bunch on TWD running for much longer (just because we can). Not really necessary but the plots will be nicer to look at.

[geraintpalmer]

Varying n1

[drvinceknight]

Longer runs.

On Thu, 30 Apr 2015 17:14 Geraint Palmer notifications@github.com wrote:

Varying n1 [image: plot_vary_n1] https://cloud.githubusercontent.com/assets/9679702/7417194/4d9dbb86-ef5c-11e4-9f76-4ea4862c9af4.png

— Reply to this email directly or view it on GitHub https://github.com/geraintpalmer/DetectingDeadlockInQingNetworkSimulation/issues/14#issuecomment-97862265 .

[geraintpalmer]

Varying L2

[drvinceknight]

More iterations.

On Thu, 30 Apr 2015 20:22 Geraint Palmer notifications@github.com wrote:

Varying L2 [image: plot_vary_l2] https://cloud.githubusercontent.com/assets/9679702/7420707/970277b6-ef76-11e4-9f57-fe170c8e0fd0.png

— Reply to this email directly or view it on GitHub https://github.com/geraintpalmer/DetectingDeadlockInQingNetworkSimulation/issues/14#issuecomment-97935218 .

[geraintpalmer]

Now for the 2 node network, simulating for 10000 iterations: Base parameters: L1 = 4.0 L2 = 5.0 mu1 = 10.0 mu2 = 8.0 r12 = 0.25 r21 = 0.15 n1 = 3 n2 = 2

• [x] L1, 2.5-14 in steps of 0.5
• [x] L2, 2.5-14 in steps of 0.5
• [ ] mu1, 2.5-14 in steps of 0.5
• [ ] mu2, 2.5-14 in steps of 0.5
• [x] r12, 0.1-1.0 in steps of 0.05
• [ ] r21, 0.1-1.0 in steps of 0.05
• [x] n1, 0-6 in steps of 1
• [x] n2, 0-6 in steps of 1

[geraintpalmer]

varying L1 10000 itrs

[geraintpalmer]

varying n1, 10000 itrs

[drvinceknight]

Are these looking any better or are they the same as for 1000 iterations? Looks similar to me...

On Wed, 6 May 2015 17:58 Geraint Palmer notifications@github.com wrote:

varying n1, 10000 itrs [image: plot_vary_n1] https://cloud.githubusercontent.com/assets/9679702/7498535/7373247e-f419-11e4-9003-6e9ca64e88ca.png

— Reply to this email directly or view it on GitHub https://github.com/geraintpalmer/DetectingDeadlockInQingNetworkSimulation/issues/14#issuecomment-99537448 .

[geraintpalmer]

Ie they look very similar to me

[drvinceknight]

Cool.

On Wed, 6 May 2015 18:01 Geraint Palmer notifications@github.com wrote:

Ie they look very similar to me

— Reply to this email directly or view it on GitHub https://github.com/geraintpalmer/DetectingDeadlockInQingNetworkSimulation/issues/14#issuecomment-99538359 .

[geraintpalmer]

Varying L2

[geraintpalmer]

Varying n2

[geraintpalmer]

Varying r12

[drvinceknight]

Is the calculation of the mean of the simulation taking account of the removed outliers? If not perhaps throw it back in (this is a pie in the sky suggestion).

On Thu, May 7, 2015 at 3:08 PM Geraint Palmer notifications@github.com wrote:

Varying n2 [image: plot_vary_n2] https://cloud.githubusercontent.com/assets/9679702/7517089/d7be77b0-f4ca-11e4-934b-c5ee660224b1.png

— Reply to this email directly or view it on GitHub https://github.com/geraintpalmer/DetectingDeadlockInQingNetworkSimulation/issues/14#issuecomment-99880849 .

[geraintpalmer]

No, removing outliers only removes them graphically. The red square for mean takes them into account.

[drvinceknight]

I thought so... We'll take another look at the MC.

On Thu, May 7, 2015 at 3:17 PM Geraint Palmer notifications@github.com wrote:

No, removing outliers only removes them graphically. The red square for mean takes them into account.

— Reply to this email directly or view it on GitHub https://github.com/geraintpalmer/DetectingDeadlockInQingNetworkSimulation/issues/14#issuecomment-99882802 .

[geraintpalmer]

varying r21