Ventana software support forum

Right hand and left hand unit of replicator dynamics equation do not match

replicator dynamic equation：dy/dt＝y(1-y)(URC-URN)
The left hand：percent/Month

The right hand：CNY

Although Units equivalents [percent/Month,CNY] were set, this problem still exist.

Thank you for your answer

Can you upload the model?

And what units are "CNY"? I've never heard of this as a unit/dimension.

Thank you for your reply.
I have been uploaded the model.

The unit "CNY" is the chinese currency unit.

First of all, remove the units equivalence "percent/Month,CNY". CNY is currency, percent/month is definitely not. You use units equiveillances for things like "month,months" or "person,persons,people,human".

Looking at your model, start by putting in units for all of the parameters. These units should match reality, if something is an interest rate, it should be something like percent/year or percent/month etc.

I've got no idea what all the parameters refer to in this model so cannot really offer any more advice.

Thank you for your help.

I want to set the rate(variable)'s unit in eqution: dx/dt＝x'(t)＝a×x(t)，a is const.

But, units do not match(left and right hand).

x(t) represent the probability of selecting one strategy, i set the unit: Dmnl,
but the unit of [dx/dt] must be the 1/Month(or other time type), so i can not match the units between the 2 sides.

Could I explain my question clearly?

Wish for your help.
Thx...

Units should always be very straightforward. Always enter what the parameter would be measured in in the real world.

Can you upload the model? It's difficult to offer advice with so little to work on.

I'd suggest that the probability is correct, that should be dimensionless. Cannot really offer any more advice without seeing the model.

I have been uploading my model.
I am waiting for your help.
Thx

Administrator wrote: ↑Tue Apr 18, 2023 10:32 am Units should always be very straightforward. Always enter what the parameter would be measured in in the real world.

Can you upload the model? It's difficult to offer advice with so little to work on.

I'd suggest that the probability is correct, that should be dimensionless. Cannot really offer any more advice without seeing the model.

The unitrate was set as Tolerances for Arithmetic Sequences, i.e. dx/dt=Δx, the unit of left is 1/time, but the unit of right is 1.
How can I match the units?

Thx

Fist thing, δx should have units of DMNL/month as it's a rate of change.

I cannot really offer any more assistance as the model is not documented, I've got no idea what things like "L1" are (or if the units for it are even correct).

Administrator wrote: ↑Thu Apr 20, 2023 1:41 pm Fist thing, δx should have units of DMNL/month as it's a rate of change.

I cannot really offer any more assistance as the model is not documented, I've got no idea what things like "L1" are (or if the units for it are even correct).

When I simulate， there will be a variable out of range fault.
How can I debug?

I am waiting for your help.
Thx

lagraberry wrote: ↑Mon Apr 24, 2023 6:39 am

Administrator wrote: ↑Thu Apr 20, 2023 1:41 pm Fist thing, δx should have units of DMNL/month as it's a rate of change.

I cannot really offer any more assistance as the model is not documented, I've got no idea what things like "L1" are (or if the units for it are even correct).

When I simulate， there will be a variable out of range fault.
How can I debug?

I am waiting for your help.
Thx

This is the next variable definition.

By multiplying the right-hand side of the equation by this conversion factor, you can ensure that the units match. The modified equation would then be:

dy/dt = y(1-y) * conversion_factor * (URC - URN)