
Fast Grouping and Ordering
fast-grouping-ordering.Rdcollapse provides the following functions to efficiently group and order data:
radixorder, provides fast radix-ordering through direct access to the methodorder(..., method = "radix"), as well as the possibility to return some attributes very useful for grouping data and finding unique elements.radixordervexists as a programmers alternative. The functionroworder(v)efficiently reorders a data frame based on an ordering computed byradixorderv.groupprovides fast grouping in first-appearance order of rows, based on a hashing algorithm in C. Objects have class 'qG', see below.GRPcreates collapse grouping objects of class 'GRP' based onradixordervorgroup. 'GRP' objects form the central building block for grouped operations and programming in collapse and are very efficient inputs to all collapse functions supporting grouped operations.fgroup_byprovides a fast replacement fordplyr::group_by, creating a grouped data frame (or data.table / tibble etc.) with a 'GRP' object attached. This grouped frame can be used for grouped operations using collapse's fast functions.fmatchis a fast alternative tomatch, which also supports matching of data frame rows.funiqueis a faster version ofunique. The data frame method also allows selecting unique rows according to a subset of the columns.fnuniqueefficiently calculates the number of unique values/rows.fduplicatedis a fast alternative toduplicated.any_duplicatedis a simpler and faster alternative toanyDuplicated.fcountcomputes group counts based on a subset of columns in the data, and is a fast replacement fordplyr::count.fcountvis a programmers version of the function.qF, shorthand for 'quick-factor' implements very fast factor generation from atomic vectors using either radix orderingmethod = "radix"or hashingmethod = "hash". Factors can also be used for efficient grouped programming with collapse functions, especially if they are generated usingqF(x, na.exclude = FALSE)which assigns a level to missing values and attaches a class 'na.included' ensuring that no additional missing value checks are executed by collapse functions.qG, shorthand for 'quick-group', generates a kind of factor-light without the levels attribute but instead an attribute providing the number of levels. Optionally the levels / groups can be attached, but without converting them to character. Objects have a class 'qG', which is also recognized in the collapse ecosystem.fdroplevelsis a substantially faster replacement fordroplevels.finteractionis a fast alternative tointeractionimplemented as a wrapper aroundas_factor_GRP(GRP(...)). It can be used to generate a factor from multiple vectors, factors or a list of vectors / factors. Unused factor levels are always dropped.groupidis a generalization ofdata.table::rleidproviding a run-length type group-id from atomic vectors. It is generalization as it also supports passing an ordering vector and skipping missing values. For exampleqFandqGwithmethod = "radix"are essentially implemented usinggroupid(x, radixorder(x)).seqidis a specialized function which creates a group-id from sequences of integer values. For any regular panel datasetgroupid(id, order(id, time))andseqid(time, order(id, time))provide the same id variable.seqidis especially useful for identifying discontinuities in time-sequences.timeidis a specialized function to convert integer or double vectors representing time (such as 'Date', 'POSIXct' etc.) to factor or 'qG' object based on the greatest common divisor of elements (thus preserving gaps in time intervals).
Table of Functions
| Function / S3 Generic | Methods | Description | ||
radixorder(v) | No methods, for data frames and vectors | Radix-based ordering + grouping information | ||
roworder(v) | No methods, for data frames incl. pdata.frame | Row sorting/reordering | ||
group | No methods, for data frames and vectors | Hash-based grouping + grouping information | ||
GRP | default, GRP, factor, qG, grouped_df, pseries, pdata.frame | Fast grouping and a flexible grouping object | ||
fgroup_by | No methods, for data frames | Fast grouped data frame | ||
fmatch | No methods, for vectors and data frames | Fast matching | ||
funique, fnunique, fduplicated, any_duplicated | default, data.frame, sf, pseries, pdata.frame, list | Fast (number of) unique values/rows | ||
fcount(v) | Internal generic, supports vectors, matrices, data.frames, lists, grouped_df and pdata.frame | Fast group counts | ||
qF | No methods, for vectors | Quick factor generation | ||
qG | No methods, for vectors | Quick grouping of vectors and a 'factor-light' class | ||
fdroplevels | factor, data.frame, list | Fast removal of unused factor levels | ||
finteraction | No methods, for data frames and vectors | Fast interactions | ||
groupid | No methods, for vectors | Run-length type group-id | ||
seqid | No methods, for integer vectors | Run-length type integer sequence-id | ||
timeid | No methods, for integer or double vectors | Integer-id from time/date sequences |