SCIP_COL

Type Alias SCIP_COL 

Source
pub type SCIP_COL = SCIP_Col;
Expand description

column of an LP

  • \ref PublicColumnMethods “List of all available methods”

Aliased Type§

#[repr(C)]
pub struct SCIP_COL {
Show 43 fields pub obj: f64, pub lb: f64, pub ub: f64, pub unchangedobj: f64, pub lazylb: f64, pub lazyub: f64, pub flushedobj: f64, pub flushedlb: f64, pub flushedub: f64, pub primsol: f64, pub redcost: f64, pub farkascoef: f64, pub minprimsol: f64, pub maxprimsol: f64, pub sbdown: f64, pub sbup: f64, pub sbsolval: f64, pub sblpobjval: f64, pub sbnode: i64, pub obsoletenode: i64, pub storedsolvals: *mut SCIP_ColSolVals, pub var: *mut SCIP_Var, pub rows: *mut *mut SCIP_Row, pub vals: *mut f64, pub validredcostlp: i64, pub validfarkaslp: i64, pub validsblp: i64, pub linkpos: *mut i32, pub index: i32, pub size: i32, pub len: i32, pub nlprows: i32, pub nunlinked: i32, pub lppos: i32, pub lpipos: i32, pub lpdepth: i32, pub sbitlim: i32, pub nsbcalls: i32, pub age: i32, pub var_probindex: i32, pub _bitfield_align_1: [u8; 0], pub _bitfield_1: __BindgenBitfieldUnit<[u8; 2]>, pub __bindgen_padding_0: [u16; 3],
}

Fields§

§obj: f64

< current objective value of column in LP (might be changed in diving or probing)

§lb: f64

< current lower bound of column in LP

§ub: f64

< current upper bound of column in LP

§unchangedobj: f64

< unchanged objective value of column (ignoring diving or probing changes)

§lazylb: f64

< lazy lower bound of the column; if the current lower bound is not greater than the lazy lower bound, then the lower bound has not to be added to the LP

§lazyub: f64

< lazy upper bound of the column; if the current upper bound is not smaller than the lazy upper bound, then the upper bound has not to be added to the LP

§flushedobj: f64

< objective value of column already flushed to the LP solver

§flushedlb: f64

< lower bound of column already flushed to the LP solver

§flushedub: f64

< upper bound of column already flushed to the LP solver

§primsol: f64

< primal solution value in LP, is 0 if col is not in LP

§redcost: f64

< reduced cost value in LP, or SCIP_INVALID if not yet calculated

§farkascoef: f64

< coefficient in dual Farkas infeasibility proof (== dualfarkas^T A_c)

§minprimsol: f64

< minimal LP solution value, this column ever assumed

§maxprimsol: f64

< maximal LP solution value, this column ever assumed

§sbdown: f64

< strong branching information for downwards branching

§sbup: f64

< strong branching information for upwards branching

§sbsolval: f64

< LP solution value of column at last strong branching call

§sblpobjval: f64

< LP objective value at last strong branching call on the column

§sbnode: i64

< node number of the last strong branching call on this column

§obsoletenode: i64

< last node where this column was removed due to aging

§storedsolvals: *mut SCIP_ColSolVals

< values stored before entering diving or probing mode

§var: *mut SCIP_Var

< variable, this column represents; there cannot be a column without variable

§rows: *mut *mut SCIP_Row

< rows of column entries, that may have a nonzero dual solution value

§vals: *mut f64

< coefficients of column entries

§validredcostlp: i64

< LP number for which reduced cost value is valid

§validfarkaslp: i64

< LP number for which Farkas coefficient is valid

§validsblp: i64

< LP number for which strong branching values are valid

§linkpos: *mut i32

< position of col in col vector of the row, or -1 if not yet linked

§index: i32

< consecutively numbered column identifier

§size: i32

< size of the row- and val-arrays

§len: i32

< number of nonzeros in column

§nlprows: i32

< number of linked rows in column, that belong to the current LP

§nunlinked: i32

< number of column entries, where the rows don’t know about the column

§lppos: i32

< column position number in current LP, or -1 if not in current LP

§lpipos: i32

< column position number in LP solver, or -1 if not in LP solver

§lpdepth: i32

< depth level at which column entered the LP, or -1 if not in current LP

§sbitlim: i32

< strong branching iteration limit used to get strong branching values, or -1

§nsbcalls: i32

< number of times, strong branching was applied on the column

§age: i32

< number of successive times this variable was in LP and was 0.0 in solution

§var_probindex: i32

< copy of var->probindex for avoiding expensive dereferencing

§_bitfield_align_1: [u8; 0]§_bitfield_1: __BindgenBitfieldUnit<[u8; 2]>§__bindgen_padding_0: [u16; 3]