ramsort {bit} | R Documentation |
These are generic stubs for low-level sorting and ordering methods implemented in packages
'bit64' and 'ff'.
The ..sortorder
methods do sorting and ordering at once, which requires more RAM than ordering but is (almost) as fast as as sorting.
ramsort(x, ...) ramorder(x, i, ...) ramsortorder(x, i, ...) mergesort(x, ...) mergeorder(x, i, ...) mergesortorder(x, i, ...) quicksort(x, ...) quickorder(x, i, ...) quicksortorder(x, i, ...) shellsort(x, ...) shellorder(x, i, ...) shellsortorder(x, i, ...) radixsort(x, ...) radixorder(x, i, ...) radixsortorder(x, i, ...) keysort(x, ...) keyorder(x, i, ...) keysortorder(x, i, ...)
x |
a vector to be sorted by |
i |
integer positions to be modified by |
... |
further arguments to the sorting methods |
The sort
generics do sort their argument 'x', some methods need temporary RAM of the same size as 'x'.
The order
generics do order their argument 'i' leaving 'x' as it was,
some methods need temporary RAM of the same size as 'i'.
The sortorder
generics do sort their argument 'x' and order their argument 'i',
this way of ordering is much faster at the price of requiring temporary RAM for both,
'x' and 'i', if the method requires temporary RAM.
The ram
generics are high-level functions containing an optimizer that chooses the 'best' algorithms given some context.
These functions return the number of NAs
found or assumed during sorting
generic | ff | bit64 |
ramsort | ramsort.default | ramsort.integer64 |
shellsort | shellsort.default | shellsort.integer64 |
quicksort | quicksort.integer64 |
|
mergesort | mergesort.default | mergesort.integer64 |
radixsort | radixsort.default | radixsort.integer64 |
keysort | keysort.default | |
generic | ff | bit64 |
ramorder | ramorder.default | ramorder.integer64 |
shellorder | shellorder.default | shellorder.integer64 |
quickorder | quickorder.integer64 |
|
mergeorder | mergeorder.default | mergeorder.integer64 |
radixorder | radixorder.default | radixorder.integer64 |
keyorder | keyorder.default | |
generic | ff | bit64 |
ramsortorder | ramsortorder.integer64 |
|
shellsortorder | shellsortorder.integer64 |
|
quicksortorder | quicksortorder.integer64 |
|
mergesortorder | mergesortorder.integer64 |
|
radixsortorder | radixsortorder.integer64 |
|
keysortorder | ||
Note that these methods purposely violate the functional programming paradigm: they are called for the side-effect of changing some of their arguments.
The rationale behind this is that sorting is very RAM-intensive and in certain
situations we might not want to allocate additional memory if not necessary to do so.
The sort
-methods change x
, the order
-methods change i
, and the sortoder
-methods change both x
and i
You as the user are responsible to create copies of the input data 'x' and 'i'
if you need non-modified versions.
Jens Oehlschlägel <Jens.Oehlschlaegel@truecluster.com>