search

Helfima / helmod

Factorio Mod
149 stars 64 forks source link

Output quantity is wrong #414

Closed KiwiHawk closed 5 months ago

KiwiHawk commented 1 year ago

Mod list

In the following screenshot, output of Omnic Waste should be 0, not 170.3

image

Import string H4sIAAAAAAAA/91Xz2+jPBD9VyrODjK/qSruK+2t11UVOeBkrRqMwDQbVfzvO8YYTCHdNt/la5GlYGfm2fPmeTIpxB0XOeF3++yVFZlTioLyfeAgca5okzk/ 2Zn9IOdnBx3A7rnNXofPvafNx4mDKlLSzBFlxfLdmbSS7hqaX3LOqtMGVi66SmYeqsUZvsAIrBpaMFpJ2OCXDeM8Za8rbAfJSw1LR96xwqBh1EoiaYZ7WCDc uKl3Y88kLSdzF+MkjUPPm0YQjRA+kgx8vR7NZ5G0WZ9FLW6dxQ/dwA+n4d+HK2QFAQsW4BzX8pyh78XWiNK3WD2qG1F0+ZK8Tx3YkOf3aDAyftc9PKzSVnf6 TXt7G7x9Kodegl0/wfOIR+RgjhV0xWoKoT6OGnwE+bGqoH9g/3kbMNPx7gZuG5JLJqodMftqmJllP3nzGFLBC4CP4C+aiwnk0DWVDS0mkkDDTF5mXC9Ilw+i 4Hm6KNqwYq6mtIANOCuZygNcwI6r8HpEj0c6JFXb4OlEL7ADTHNRtV1ZDyeELwXnnX6HG0DqW9ysiYuTaDzqXgoJF8qyNSsu9nDoxdYToWNH+XjtenSgBLab WBtmmyo3rITpwEouyoPIQlTTZm+ox25kz/cqDknAOdgkb3H0ftTPLRH16NEfleYbpeH3C962xEIc2SPxbInhdyRW/+bidE1ga1RD5f9dYPcLnuP4I9kJkkUx TMMvr7dVRKC3aNRbZPTmv6lsVj1f1rfimvhwEHmLES/U56ZJGlgj9P9DyUui2B7freQFywDTLy/BVUTwM2vqXFdBSXsGUmXTUdQK/gLdwfC+1Nm42QZ6DLSG ETbD9xDkoiRKWYeO8QIqJsSjQ2R06GE2tPb0bx3a7PqqB1mV0C2QRXG1IVSzodOnut6PZhWc5pQAwGVK5pHwlg7Mbja7Y7+meY/RzY3m2wZTAybolkbTarpa 0TW51tmYahBCq+6Ng917N4jmJkz/KWDqimt1qPfIalLHFnNZ2t5vNa82lhokWP8ef7TBnGKcVPktpYisuzYWHrwsfqt72qOTsiQQFyCQPO/KjquZ2dta2j4A BA4VgzS7mlSUzzHbi1cdJSXljlYnVln/IBar20H3Dw2VQOrd/oFWxV8kQMsp4Q4AAA==
KiwiHawk commented 1 year ago

I don't think matrix solver should be able to calculate this. The percentage for the first recipe can't be set as there is no demand for Omnic Water at that stage.

I would expect it to look like this: image

KiwiHawk commented 1 year ago

Problem is in SolverSimplex:lineCompute. Production and Coefficient are both being set to 0 as Z demand is 0: https://github.com/Helfima/helmod/blob/2123046d00ccd1141448a2ec75dc99168808cbc5/math/SolverSimplex.lua#L195

Because of this, Z running total is not updated: https://github.com/Helfima/helmod/blob/2123046d00ccd1141448a2ec75dc99168808cbc5/math/SolverSimplex.lua#L213

image

KiwiHawk commented 1 year ago

Maybe the recipe lines should automatically be reordered? Only when using Matrix Solver. And only for lines with no prior demand.

Also delete duplicate recipes.

Helfima commented 5 months ago

same issue Matrix solver with omni fuel #415 the input must be the primal recipe we can do better image