This function requires the bigdata library section to be loaded.
bd.split.by.window(data, window, offset=0, drop.incomplete=F,
bigdata=is(x,"bdFrame"))
bdFrame or
data.frame.
offset=window, so each block directly follows
the previous one.
If
offset is greater than
window,
some rows will be skipped between blocks.
TRUE, this will only process blocks with
window rows.
If this is
FALSE, blocks at the end of the data set will be
processed, even if they have less than
window rows.
TRUE, returns a list of
bdFrame objects.
If
FALSE, this returns a list of
data.frame objects.
The default uses the type of
x to determine which type of objects to return.
bigdata is
TRUE,
the list elements will be
bdFrame objects;
otherwise, they will be
data.frame objects.
This function divides the input data into blocks defined by a moving window over the data rows, and returns a list of all of these blocks.
If
bigdata is
FALSE, the output list elements will be
data.frame objects.
In this case, if all of the data is too large to fit in memory, an error will occur.
## Divide fuel.frame into a list of 5-row data.frames bd.split.by.window(fuel.frame, 5)