Need help understanding new heat exchangers

Quote

For this reason, I hereby present you: HeatSwitches (commonly known as HD or HeatDissipator, HeatDistributer and Strange-Thing-Which-Can-Magically-Alter-Temeperatures).
The standard HeatSwitch can store 2500 heat, has a 'sideTransfer rate' of 12 and a 'coreTransfer rate' of 4.
All HeatSwitches work the same way: They calculate the % of heat stored in all surrounding tiles, themselves and the reactor hull, calculate a median and then attempt to reach that median on all components. A heatSwitch will first shift around (component <-> switch) the heat of adjacent components, to a max of sideTransfer. Then he will try to balance the heat between itself and the reactor to a max of coreTransfer.
The 'Core Heat Switch' does have a sideTransfer rate of 0 (thus no heat balance between adjacent components), but a coreTransfer rate of 72, and a maxHeat of 5000.
The 'Spread Heat Switch' does not have a coreTrasnfer, but instead 36 sideTransfer, and a maxHeat of 5000.
Lastly, the 'Diamond Heat Switch' has a sideTransfer of 24 and a coreTransfer of 8, and a maxHeat of 10000.
Opposed to the old HD's, the switches do NOT dissipate heat, have a LOW heat storage and do go by %, not my static values. F.e. you have a core heat switch (5000 max) and a reactor with some plating (20000 max). The system has a total of 5000 heat. The switch will balance 1000 heat to itself and 4000 to the reactor, resulting in 20% heat for itself and the reactor.

Display More

I'm not understanding how the math is done above.
Start:
Reactor Hul = 5,000 Heat
Max Reactor Heat = 20,000 Heat
Core Heat Exchanger = 0 Heat

How exactly is the median heat value calculated?
The median value given the numbers {0, 5000} is 2,500.
The median value given the numbers {0, 5000, 20000} is 5,000.

The Hull is at 25% of its heat capacity.
The Core Heat Exchanger is at 0% of its heat capacity.

I'm not understanding how the Core Heat Exchanger finally determines that it should 1,000 heat in itself and 4,000 heat in the reactor.
I do see that when the Core Heat Exchanger is at 1,000 of 5,000 heat that's at 20% capacity, and when the reactor hull is at 4,000 of 20,000 heat it is also at 20% capacity.

So is it taking that initial 5,000 heat and trying to distribute it in such a way that the % capacity of each component it affects (itself & hull) are the same? That seems like the answer and makes sense after I've worked through typing all of this out. But how does it mathematically find that value?
Core Heat Exchanger
X/5000 = Y Where X is the amount of heat to store inside of the Core Heat Exchanger, and Y is the % of heat stored in the component
Z/20000 = Y Where Z is the amount of heat to store inside the Hull

Is that what it's doing?

X/5000 = Z/20000
4 = Z/X
X + Z = 5000
Z = 5000 - X
4 = (5000 - X)X
4X = 5000 - X
4X + X = 5000
X(4 + 1) = 5000
5X = 5000
X = 1000
Z = 4000

Ok... so I think I've got that example figured out. Substitute an Advanced Heat Exchanger for the Core Heat Exchanger and also toss in a 60k Coolant Cell and I think I'll wait for the reactor planner to get updated.

Yeah that's a lot simpler.

Based upon the original design wording, the same works for component <-> component heat as well I'm assuming.

Start:
10k coolant cell with 9k heat <-> component heat exchanger <-> overclocked heat vent

The total capacity of those 3 items = 10k + 5k + 1k = 16k
The first tick for the exchanger would add that up, and then take 9k/16k = 0.5625
So it would pull 36 heat out of the 10k coolant cell (which is max transfer for its side) and then put 36 heat into the overclocked heat vent?

Or maybe a simpler approach, if there is only the 10k coolant cell with 9k heat and the component heat exchanger. The "fillgrade" is 0.6. So each tick the exchanger will pull 36 heat from the coolant cell and store it in itself until the coolant cell is at 6k heat and the exchanger has 3k heat.

If the Overclocked Heat Vent is given 36 heat, it will only dissipate 20 of that heat, leaving 16 heat in it.
So every tick the exchanger would be giving the vent more heat than it could vent... however it would also be getting rid of 20 heat each time as well.

Since the vent can only store 1k heat, the exchanger should prevent the vent from burning itself out, because as the vent approached its 1k heat max, the % heat stored in the vent would be greater than that "fillgrade" so the exchanger would actually start to pull excess heat away from the vent, and possibly store it either in itself or shuffle it back to the coolant cell depending upon how much heat was currently stored in each one.

I think I've got a pretty good handle on how it all works now. Showing me that better formula really helped out a lot.