<s>
HiGHS	O
is	O
open-source	O
software	O
to	O
solve	O
linear	B-Algorithm
programming	I-Algorithm
(	O
LP	O
)	O
,	O
mixed-integer	B-Algorithm
programming	I-Algorithm
(	O
MIP	O
)	O
,	O
and	O
convex	O
quadratic	B-Algorithm
programming	I-Algorithm
(	O
QP	O
)	O
models	O
.	O
</s>
<s>
Written	O
in	O
C++	B-Language
and	O
published	O
under	O
an	O
MIT	B-License
license	I-License
,	O
HiGHS	O
provides	O
programming	B-Application
interfaces	I-Application
to	O
C	B-Language
,	O
Python	B-Language
,	O
Julia	B-Application
,	O
Rust	B-Application
,	O
JavaScript	B-Language
,	O
Fortran	B-Application
,	O
and	O
C#	B-Application
.	O
</s>
<s>
Aconvenient	O
thin	O
wrapper	O
to	O
Python	B-Language
is	O
available	O
via	O
the	O
PyPI	B-Language
package	O
.	O
</s>
<s>
HiGHS	O
is	O
designed	O
to	O
solve	O
large-scale	O
models	O
and	O
exploits	O
problem	B-Algorithm
sparsity	I-Algorithm
.	O
</s>
<s>
Its	O
origins	O
can	O
be	O
traced	O
back	O
to	O
late	O
2016	O
,	O
when	O
Ivet	O
Galabova	O
combined	O
her	O
LP	O
presolve	O
with	O
Julian	O
Hall	O
's	O
simplex	O
crash	O
procedure	O
and	O
Huangfu	O
Qi	O
's	O
dual	O
simplex	O
solver	O
to	O
solve	O
a	O
class	O
of	O
industrial	O
LP	B-Algorithm
problems	I-Algorithm
faster	O
than	O
the	O
best	O
open-source	O
solvers	O
at	O
that	O
time	O
.	O
</s>
<s>
Since	O
then	O
,	O
a	O
C++API	O
and	O
other	O
language	O
interfaces	O
have	O
been	O
developed	O
,	O
and	O
modelling	O
utilities	O
and	O
other	O
categories	O
of	O
solver	O
have	O
been	O
added	O
.	O
</s>
<s>
HiGHS	O
has	O
implementations	O
of	O
the	O
primal	O
and	O
dual	O
revised	B-Algorithm
simplex	I-Algorithm
method	I-Algorithm
for	O
solving	O
LP	B-Algorithm
problems	I-Algorithm
,	O
based	O
on	O
techniques	O
described	O
by	O
Hall	O
and	O
McKinnon	O
(	O
2005	O
)	O
,	O
and	O
Huangfu	O
and	O
Hall	O
(	O
2015	O
,	O
2018	O
)	O
.	O
</s>
<s>
These	O
include	O
the	O
exploitation	O
of	O
hyper-sparsity	O
when	O
solving	O
linear	O
systems	O
in	O
the	O
simplex	O
implementations	O
and	O
,	O
for	O
the	O
dual	O
simplex	O
solver	O
,	O
exploitation	O
of	O
multi-threading	O
.	O
</s>
<s>
HiGHS	O
has	O
an	O
interior	B-Algorithm
point	I-Algorithm
method	I-Algorithm
implementation	O
for	O
solving	O
LP	B-Algorithm
problems	I-Algorithm
,	O
based	O
on	O
techniques	O
described	O
by	O
Schork	O
and	O
Gondzio	O
(	O
2020	O
)	O
.	O
</s>
<s>
HiGHS	O
has	O
an	O
active	O
set	O
solver	O
for	O
convex	O
quadratic	B-Algorithm
programming	I-Algorithm
(	O
QP	O
)	O
problems	O
.	O
</s>
<s>
HiGHS	O
can	O
be	O
used	O
as	O
a	O
standalone	O
solver	O
library	O
in	O
bespoke	O
applications	O
,	O
but	O
numerical	B-General_Concept
computing	I-General_Concept
environments	O
,	O
optimization	O
programming	O
packages	O
,	O
and	O
domainspecific	O
numerical	B-General_Concept
analysis	I-General_Concept
projects	O
are	O
starting	O
to	O
incorporate	O
the	O
software	O
into	O
their	O
systems	O
also	O
.	O
</s>
<s>
As	O
powerful	O
opensource	O
software	O
under	O
active	O
development	O
,	O
HiGHS	O
is	O
increasingly	O
being	O
adopted	O
by	O
application	B-Application
software	I-Application
projects	O
that	O
provide	O
support	O
for	O
numerical	B-General_Concept
analysis	I-General_Concept
.	O
</s>
<s>
The	O
SciPy	B-Application
scientific	O
library	O
,	O
for	O
instance	O
,	O
uses	O
HiGHS	O
as	O
its	O
LP	O
solver	O
from	O
release1.6.0	O
and	O
the	O
HiGHS	O
MIP	O
solver	O
for	O
discrete	O
optimization	O
from	O
release1.9.0	O
.	O
</s>
<s>
As	O
well	O
as	O
offering	O
an	O
interface	O
to	O
HiGHS	O
,	O
the	O
JuMP	O
modelling	O
language	O
for	O
Julia	B-Application
also	O
describes	O
the	O
specific	O
use	O
of	O
HiGHS	O
in	O
its	O
user	O
documentation	O
.	O
</s>
