PyRO Chains Analysis

pyanodon's mods are here

Moderator: pyanodon

immortal_sniper1
Filter Inserter
Filter Inserter
Posts: 450
Joined: Sun Jun 03, 2018 8:54 am
Contact:

Re: PyRO Chains Analysis

Post by immortal_sniper1 » Tue May 14, 2019 5:48 pm

mikes61293 wrote:
Tue May 14, 2019 5:41 pm
immortal_sniper1 wrote:
Tue May 14, 2019 5:15 pm
mikes61293 wrote:
Tue May 14, 2019 4:29 pm
It looks like turning your unslimed iron into pulp and processing it into high grade iron is never worth it. I think the chain should be balanced such that converting your unslimed iron is a marginal increase in efficiency if you're smelting high grade iron (tier 3.2) and a decent increase with sinter iron (tier 4.2).
no since it would enable loops of iron and py said he doesnt want loops
not to mention that then people would get iron from oxide mostly
Iron already loops at the iron pulp 5 -> iron pulp 6 step by creating more iron pulp 2. If you remove iron pulp 2 as a byproduct from that recipe and increase the ratio of iron pulp 6 -> iron pulp 7, you'd increase the viability of that whole chain and it wouldn't loop. The whole chain would split at the classify iron ore dust step and then converge at the iron pulp 7 -> high grade iron step. If turning unslimed iron into iron pulp and processing it is always worse than just smelting the unslimed iron, why even have that chain?

Essentially I just want the highlighted section of the chart below to be worth doing.
Iron Pulp Chain 2.jpg
maybe my wording wasnt the best

in release 1 or 2 people could

iron oxide>unslimed> pulp 7> iron then take some of this iron and make iron oxide from it and get a profit of iron

Openhand
Inserter
Inserter
Posts: 28
Joined: Mon Apr 25, 2016 7:13 pm
Contact:

Re: PyRO Chains Analysis

Post by Openhand » Thu May 16, 2019 7:04 am

I installed the helmod and it is very helpful for comparing different ore processing chains, my life is easier now :)

Helmod could be better though or i'm just bad at it, like instead for aiming for an end product, how about having a defined input of like 5000 ores, what would the following picked recipes give from 5000 ore? so reversing how it works, is that possible currently?

Then you could have a true comparison between the chains i think.

Also a way to handle when a resource is used multiple times in a chain and not just once, right now it kinda hangs up and i have to add another production block for it.

Using the optimized version of helmod, the original is slow for some reason.

Also there is a bug when you check in "count by factory" or something and then remove it, the results can stay longer than wanted and it can be very wrong.

Also i don't understand how linking does anything and i like how i can setup a product block automatically to handle all the oxygen need, but why doesn't it work for other resources?

Like all the resources that doesn't get used can be summarized and handled by a recipe of your choice (if you choose to do it), like the by-products, all the different by-products should (and is i think) be visible and when you add a different production block for them, it automatically tries to use those to their full extent, instead of now how i have to manually do guess work to get the exact number of by-products to produce X plates, this guessing game is dumb.

I start by aiming for 900 plates, then it is molten metal, then maybe some dust of some sort and other by-products, then i do plates from each by-product separately by guessing the plates and looking at the input by-products and try to match them exactly to the total from my output of the original plate chain.

It should automatically try and match that number is what i'm saying, has that something to do with linking perhaps? i dunno, but for oxygen it just works, with or without linking iirc.

Blokus
Fast Inserter
Fast Inserter
Posts: 117
Joined: Tue Mar 05, 2019 2:49 pm
Contact:

Re: PyRO Chains Analysis

Post by Blokus » Thu May 16, 2019 3:38 pm

Openhand wrote:
Thu May 16, 2019 7:04 am
I installed the helmod and it is very helpful for comparing different ore processing chains, my life is easier now :)

Helmod could be better though or i'm just bad at it, like instead for aiming for an end product, how about having a defined input of like 5000 ores, what would the following picked recipes give from 5000 ore? so reversing how it works, is that possible currently?

Then you could have a true comparison between the chains i think.

Also a way to handle when a resource is used multiple times in a chain and not just once, right now it kinda hangs up and i have to add another production block for it.

Using the optimized version of helmod, the original is slow for some reason.

Also there is a bug when you check in "count by factory" or something and then remove it, the results can stay longer than wanted and it can be very wrong.

Also i don't understand how linking does anything and i like how i can setup a product block automatically to handle all the oxygen need, but why doesn't it work for other resources?

Like all the resources that doesn't get used can be summarized and handled by a recipe of your choice (if you choose to do it), like the by-products, all the different by-products should (and is i think) be visible and when you add a different production block for them, it automatically tries to use those to their full extent, instead of now how i have to manually do guess work to get the exact number of by-products to produce X plates, this guessing game is dumb.

I start by aiming for 900 plates, then it is molten metal, then maybe some dust of some sort and other by-products, then i do plates from each by-product separately by guessing the plates and looking at the input by-products and try to match them exactly to the total from my output of the original plate chain.

It should automatically try and match that number is what i'm saying, has that something to do with linking perhaps? i dunno, but for oxygen it just works, with or without linking iirc.
Helmod is inherently output-based. This is really bad for something like PyRO because it means that Helmod is greedy about inputs. You have a recipe 2 Y -> 3 X, you ask for 15 X/s, which means you need 10 Y/s. But there are two ways to make Y. Helmod, within a single block, doesn't care. It assigns all the production of Y to the first recipe you give it for Y and thus winds up assigning no production at all to the other recipe for Y. In PyRO this means you wind up with very large excesses of intermediates and it also predicts that your blocks need to be way bigger than they really need to be in order to achieve your targets.

The solutions that I know of:
  • Manually tweak percentages on the left side of the block screen so that the intermediates zero out
  • Whenever you have a product that gets made in two different places, give each of those chains its own block, and then balance those against each other when you assemble the blocks into a line
  • Don't use Helmod for this sort of thing, except as a simple calculator/data source. This kind of thing is straightforward to do with other tools like a spreadsheet, but the pain is getting the raw data into the other tools.
Technically the matrix solver can sometimes handle this sort of thing, but it rarely seems to work for me.

Openhand
Inserter
Inserter
Posts: 28
Joined: Mon Apr 25, 2016 7:13 pm
Contact:

Re: PyRO Chains Analysis

Post by Openhand » Sun May 19, 2019 10:08 pm

I'm solving it manually by adding the by-product processing to another product line, haven't used the percentages yet though.

Helmod i find useful to get a good overview of how many and what buildings i need to build to get to a result, its bad with by-products though i'm starting to want to treat by-products as something seperate from the main thing anyways, so not much of an issue anymore, there is a main production of plates then the by-products i think i'll just handle at a by-product site and use those plates (perhaps) for maybe just restocking player plate levels and not for other production, a use that isn't too taxing, and eventually use burners if the warehouses get full and deal with ashes in some way, maybe make it liquid and void it or gas and vent it, if i cannot find a use that would take a long time to produce a blockage.

I'm planning on using trains to export all the by-products to a by-product exclusive storage and "shipping place / warehouse" sort of thing.

Just exporting all the stuff that isn't essential to the things i'm doing at one location i think will make everything much more streamlined and easier to be built and planned.

Like my now third or fourth power station, i'm voiding all the tar instead of converting it to electricity because if i'm not using enough electricity, the tar would build up and produce a block and eventually shut the whole base factory down, a solution to this might be circuitry and valves but i think i rather just use the excess tar for much more useful applications than just more power, tho that means i'm pretty wasteful with my power production but i quadruppled my power production recently so i'm fine anyways, and i'm interested in more ways of making power than the basic: "water -> fawogae -> coal -> coal gas -> syngas -> combustion mix -> elecricity" conversion, also convert some coal to coke which is needed for the powerhouse so there is no coke overproduction in the power line, its not possible to get a power production block anymore because i'm voiding by-products... except for ash and iron oxide which i'm planning on having on a by-product export station.

It might be beautiful having a by-product processing site, no more messy by-products all over the place, just push it all to the by-product junkyard pretty much for very useful things :)

In short: having fun times with pyanodon's and i just saw the veritasium video on the 96,000,000 black balls after the video why that video got viral and they mentioned boron which comes from borax which is used in pyanodon and i just think how great modding and pyanodon is for factorio.

My factory is now close to 140h playtime, still haven't automated circuit production and i'm getting sick of building all the buildings and circuits manually and running all over the place for plates and wood and such. (need to move my lead plate production to my new plate production block, i also produce tracks at the lead production site, and wood).

So much to do, i'm pretty convinced i could probably play 1000 hours on this playthrough thanks to pyanodon's :D

User avatar
pyanodon
Smart Inserter
Smart Inserter
Posts: 1627
Joined: Wed Apr 20, 2016 4:42 pm
Contact:

Re: PyRO Chains Analysis

Post by pyanodon » Fri May 24, 2019 3:26 pm

:shock: thank you :)
pY Coal processing mod
Discord: Pyanodon #5791

ShadowGlass
Burner Inserter
Burner Inserter
Posts: 11
Joined: Sat Aug 27, 2016 1:50 pm
Contact:

Re: PyRO Chains Analysis

Post by ShadowGlass » Sun Jul 07, 2019 4:31 pm

immortal_sniper1 wrote:
Tue May 14, 2019 5:48 pm
maybe my wording wasnt the best

in release 1 or 2 people could

iron oxide>unslimed> pulp 7> iron then take some of this iron and make iron oxide from it and get a profit of iron
This loop still exists if you directly smelt the unslimed, unless I'm really miscalculating something:

Code: Select all

100 Iron Plate => 200 Iron Oxide => 500 Iron Slime => 5 Unslimed Iron => 450 Molten Iron => 270 Iron Plate
That's a +170% gain per loop, pretty huge. You need 10 Nitrobenzene for each starting Iron Plate, so you get 1.7 Iron Plates / 10 Nitrobenzene invested.

There's also a smaller loop with the Hot Air recipe:

Code: Select all

100 Iron Plate => 200 Iron Oxide => 120 Iron Plate 
That's only +20%, so it's questionable if 10 Nitrobenzene is worth 0.2 Iron Plates, in exchange for not needing any ore.

I'd recommend just nerfing the starting recipe, which is currently 5 Iron Plate => 10 Iron Oxide. Just change it to 5 => 5. The first loop above would still be positive, but much less efficient.

immortal_sniper1
Filter Inserter
Filter Inserter
Posts: 450
Joined: Sun Jun 03, 2018 8:54 am
Contact:

Re: PyRO Chains Analysis

Post by immortal_sniper1 » Sun Jul 07, 2019 10:20 pm

ShadowGlass wrote:
Sun Jul 07, 2019 4:31 pm
immortal_sniper1 wrote:
Tue May 14, 2019 5:48 pm
maybe my wording wasnt the best

in release 1 or 2 people could

iron oxide>unslimed> pulp 7> iron then take some of this iron and make iron oxide from it and get a profit of iron
This loop still exists if you directly smelt the unslimed, unless I'm really miscalculating something:

Code: Select all

100 Iron Plate => 200 Iron Oxide => 500 Iron Slime => 5 Unslimed Iron => 450 Molten Iron => 270 Iron Plate
That's a +170% gain per loop, pretty huge. You need 10 Nitrobenzene for each starting Iron Plate, so you get 1.7 Iron Plates / 10 Nitrobenzene invested.

There's also a smaller loop with the Hot Air recipe:

Code: Select all

100 Iron Plate => 200 Iron Oxide => 120 Iron Plate 
That's only +20%, so it's questionable if 10 Nitrobenzene is worth 0.2 Iron Plates, in exchange for not needing any ore.

I'd recommend just nerfing the starting recipe, which is currently 5 Iron Plate => 10 Iron Oxide. Just change it to 5 => 5. The first loop above would still be positive, but much less efficient.
nitrobenzene is expensive

ill look deper into te loops BUT

for example iron >drill head>iron ore>iron is a loop it is positive BUT the secondary inputs are so large it is not worth it

ShadowGlass
Burner Inserter
Burner Inserter
Posts: 11
Joined: Sat Aug 27, 2016 1:50 pm
Contact:

Re: PyRO Chains Analysis

Post by ShadowGlass » Mon Jul 08, 2019 12:16 am

immortal_sniper1 wrote:
Sun Jul 07, 2019 10:20 pm
nitrobenzene is expensive

ill look deper into te loops BUT

for example iron >drill head>iron ore>iron is a loop it is positive BUT the secondary inputs are so large it is not worth it
I brought this up, because you've said earlier that the Iron Pulp 2 chain is not getting buffed, because it'd create a loop. The Unslimed Iron => Molten Iron loop is basically the same loop, but much more cheaper and more efficient. If this is not a problem, then the Iron Pulp chain shouldn't be a problem either. It's a shame that there's this nice long, intricate Iron Pulp chain, and it's never used, because the Unslimed Iron => Molten Iron direct recipe is just several times better. It might be that recipe is just too good. It's only Green Science, maybe it shouldn't be that efficient. It's producing 2.16 plate/ore, while all the other metals are 1 plate/ore or lower at this tech level.

immortal_sniper1
Filter Inserter
Filter Inserter
Posts: 450
Joined: Sun Jun 03, 2018 8:54 am
Contact:

Re: PyRO Chains Analysis

Post by immortal_sniper1 » Mon Jul 08, 2019 3:30 pm

ShadowGlass wrote:
Mon Jul 08, 2019 12:16 am
immortal_sniper1 wrote:
Sun Jul 07, 2019 10:20 pm
nitrobenzene is expensive

ill look deper into te loops BUT

for example iron >drill head>iron ore>iron is a loop it is positive BUT the secondary inputs are so large it is not worth it
I brought this up, because you've said earlier that the Iron Pulp 2 chain is not getting buffed, because it'd create a loop. The Unslimed Iron => Molten Iron loop is basically the same loop, but much more cheaper and more efficient. If this is not a problem, then the Iron Pulp chain shouldn't be a problem either. It's a shame that there's this nice long, intricate Iron Pulp chain, and it's never used, because the Unslimed Iron => Molten Iron direct recipe is just several times better. It might be that recipe is just too good. It's only Green Science, maybe it shouldn't be that efficient. It's producing 2.16 plate/ore, while all the other metals are 1 plate/ore or lower at this tech level.
Unslimed Iron => Molten is the way to go since the other was nerfed for good reason
people started to sort sand to get Fe2O3 and get lots of iron from that and then loop the iron in iron STICK>stone>grave>sand and also get tons of crushed quartz in the process so no one would need to mine those 2

iron>iron oxide is expensive to the point i think it is not worth it nitrobenzenre needs benzene ammonia and H2SO4 ?
all can be made from tholens BUT do you think the vast power needed and the vast areas needed for the build are worth it ?
for small bases u might make small infinite builds and power it all by solar but for larger or faster bases the UPS cost would be great

mikes61293
Burner Inserter
Burner Inserter
Posts: 13
Joined: Fri Oct 12, 2018 12:16 am
Contact:

Re: PyRO Chains Analysis

Post by mikes61293 » Wed Jul 10, 2019 2:36 pm

ShadowGlass wrote:
Mon Jul 08, 2019 12:16 am
immortal_sniper1 wrote:
Sun Jul 07, 2019 10:20 pm
nitrobenzene is expensive

ill look deper into te loops BUT

for example iron >drill head>iron ore>iron is a loop it is positive BUT the secondary inputs are so large it is not worth it
I brought this up, because you've said earlier that the Iron Pulp 2 chain is not getting buffed, because it'd create a loop. The Unslimed Iron => Molten Iron loop is basically the same loop, but much more cheaper and more efficient. If this is not a problem, then the Iron Pulp chain shouldn't be a problem either. It's a shame that there's this nice long, intricate Iron Pulp chain, and it's never used, because the Unslimed Iron => Molten Iron direct recipe is just several times better. It might be that recipe is just too good. It's only Green Science, maybe it shouldn't be that efficient. It's producing 2.16 plate/ore, while all the other metals are 1 plate/ore or lower at this tech level.
I brought this up a few months ago in this same thread. I agree that unslimed -> Molten should be nerfed, there is no reason that the simple chain for unslimed iron should be better than the more complex chain available much later in the game. The problem is with the iron oxide conversion rate so that should be be nerfed to get rid of the iron loops.

Post Reply

Return to “PyMods”

Who is online

Users browsing this forum: katalex