Select features from a dfm or fcm

This function selects or removes features from a dfm or fcm, based on feature name matches with pattern. The most common usages are to eliminate features from a dfm already constructed, such as stopwords, or to select only terms of interest from a dictionary.

dfm_select(x, pattern = NULL, selection = c("keep", "remove"),
  valuetype = c("glob", "regex", "fixed"), case_insensitive = TRUE,
  min_nchar = 1L, max_nchar = 79L,
  verbose = quanteda_options("verbose"))

dfm_remove(x, ...)

dfm_keep(x, ...)

fcm_select(x, pattern = NULL, selection = c("keep", "remove"),
  valuetype = c("glob", "regex", "fixed"), case_insensitive = TRUE,
  verbose = quanteda_options("verbose"), ...)

fcm_remove(x, pattern = NULL, ...)

fcm_keep(x, pattern = NULL, ...)

Arguments

x	the dfm or fcm object whose features will be selected
pattern	a character vector, list of character vectors, dictionary, collocations, or dfm. See pattern for details.
selection	whether to keep or remove the features
valuetype	the type of pattern matching: "glob" for "glob"-style wildcard expressions; "regex" for regular expressions; or "fixed" for exact matching. See valuetype for details. For dfm_select, pattern may also be a dfm; see Value below.
case_insensitive	ignore the case of dictionary values if TRUE
min_nchar, max_nchar	numerics specifying the minimum and maximum length in characters for features to be removed or kept; defaults are 1 and 79. (Set max_nchar to NULL for no upper limit.) These are applied after (and hence, in addition to) any selection based on pattern matches.
verbose	if TRUE print message about how many pattern were removed
...	used only for passing arguments from dfm_remove or dfm_keep to dfm_select. Cannot include selection.

Value

A dfm or fcm object, after the feature selection has been applied.

When pattern is a dfm object and selection = "keep", then the returned object will be identical in its feature set to the dfm supplied as the pattern argument. This means that any features in x not in the dfm provided as pattern will be discarded, and that any features in found in the dfm supplied as pattern but not found in x will be added with all zero counts. Because selecting on a dfm is designed to produce a selected dfm with an exact feature match, when pattern is a dfm object, then the following settings are always used: case_insensitive = FALSE, and valuetype = "fixed".

Selecting on a dfm is useful when you have trained a model on one dfm, and need to project this onto a test set whose features must be identical. It is also used in bootstrap_dfm. See examples.

When pattern is a dfm object and selection = "keep", the returned object will simply be the dfm without the featnames matching those of the selection dfm.

Details

dfm_remove and fcm_remove are simply a convenience wrappers to calling dfm_select and fcm_select with selection = "remove".

dfm_keep and fcm_keep are simply a convenience wrappers to calling dfm_select and fcm_select with selection = "keep".

Note

This function selects features based on their labels. To select features based on the values of the document-feature matrix, use dfm_trim.

Examples

my_dfm <- dfm(c("My Christmas was ruined by your opposition tax plan.",
               "Does the United_States or Sweden have more progressive taxation?"),
             tolower = FALSE, verbose = FALSE)
my_dict <- dictionary(list(countries = c("United_States", "Sweden", "France"),
                          wordsEndingInY = c("by", "my"),
                          notintext = "blahblah"))
dfm_select(my_dfm, my_dict)
#> Document-feature matrix of: 2 documents, 4 features (50% sparse).
#> 2 x 4 sparse Matrix of class "dfm"
#>        features
#> docs    My by United_States Sweden
#>   text1  1  1             0      0
#>   text2  0  0             1      1
dfm_select(my_dfm, my_dict, case_insensitive = FALSE)
#> Document-feature matrix of: 2 documents, 1 feature (50% sparse).
#> 2 x 1 sparse Matrix of class "dfm"
#>        features
#> docs    by
#>   text1  1
#>   text2  0
dfm_select(my_dfm, c("s$", ".y"), selection = "keep", valuetype = "regex")
#> Document-feature matrix of: 2 documents, 6 features (50% sparse).
#> 2 x 6 sparse Matrix of class "dfm"
#>        features
#> docs    My Christmas was by Does United_States
#>   text1  1         1   1  1    0             0
#>   text2  0         0   0  0    1             1
dfm_select(my_dfm, c("s$", ".y"), selection = "remove", valuetype = "regex")
#> Document-feature matrix of: 2 documents, 14 features (50% sparse).
#> 2 x 14 sparse Matrix of class "dfm"
#>        features
#> docs    ruined your opposition tax plan . the or Sweden have more progressive
#>   text1      1    1          1   1    1 1   0  0      0    0    0           0
#>   text2      0    0          0   0    0 0   1  1      1    1    1           1
#>        features
#> docs    taxation ?
#>   text1        0 0
#>   text2        1 1
dfm_select(my_dfm, stopwords("english"), selection = "keep", valuetype = "fixed")
#> Document-feature matrix of: 2 documents, 9 features (50% sparse).
#> 2 x 9 sparse Matrix of class "dfm"
#>        features
#> docs    My was by your Does the or have more
#>   text1  1   1  1    1    0   0  0    0    0
#>   text2  0   0  0    0    1   1  1    1    1
dfm_select(my_dfm, stopwords("english"), selection = "remove", valuetype = "fixed")
#> Document-feature matrix of: 2 documents, 11 features (50% sparse).
#> 2 x 11 sparse Matrix of class "dfm"
#>        features
#> docs    Christmas ruined opposition tax plan . United_States Sweden progressive
#>   text1         1      1          1   1    1 1             0      0           0
#>   text2         0      0          0   0    0 0             1      1           1
#>        features
#> docs    taxation ?
#>   text1        0 0
#>   text2        1 1

# select based on character length
dfm_select(my_dfm, min_nchar = 5)
#> Document-feature matrix of: 2 documents, 7 features (50% sparse).
#> 2 x 7 sparse Matrix of class "dfm"
#>        features
#> docs    Christmas ruined opposition United_States Sweden progressive taxation
#>   text1         1      1          1             0      0           0        0
#>   text2         0      0          0             1      1           1        1

# selecting on a dfm
txts <- c("This is text one", "The second text", "This is text three")
(dfm1 <- dfm(txts[1:2]))
#> Document-feature matrix of: 2 documents, 6 features (41.7% sparse).
#> 2 x 6 sparse Matrix of class "dfm"
#>        features
#> docs    this is text one the second
#>   text1    1  1    1   1   0      0
#>   text2    0  0    1   0   1      1
(dfm2 <- dfm(txts[2:3]))
#> Document-feature matrix of: 2 documents, 6 features (41.7% sparse).
#> 2 x 6 sparse Matrix of class "dfm"
#>        features
#> docs    the second text this is three
#>   text1   1      1    1    0  0     0
#>   text2   0      0    1    1  1     1
(dfm3 <- dfm_select(dfm1, dfm2, valuetype = "fixed", verbose = TRUE))
#> kept 5 features
#> 
#> Document-feature matrix of: 2 documents, 6 features (50% sparse).
#> 2 x 6 sparse Matrix of class "dfm"
#>        features
#> docs    the second text this is three
#>   text1   0      0    1    1  1     0
#>   text2   1      1    1    0  0     0
setequal(featnames(dfm2), featnames(dfm3))
#> [1] TRUE

tmpdfm <- dfm(c("This is a document with lots of stopwords.",
                "No if, and, or but about it: lots of stopwords."),
              verbose = FALSE)
tmpdfm
#> Document-feature matrix of: 2 documents, 18 features (38.9% sparse).
#> 2 x 18 sparse Matrix of class "dfm"
#>        features
#> docs    this is a document with lots of stopwords . no if , and or but about it
#>   text1    1  1 1        1    1    1  1         1 1  0  0 0   0  0   0     0  0
#>   text2    0  0 0        0    0    1  1         1 1  1  1 2   1  1   1     1  1
#>        features
#> docs    :
#>   text1 0
#>   text2 1
dfm_remove(tmpdfm, stopwords("english"))
#> Document-feature matrix of: 2 documents, 6 features (25% sparse).
#> 2 x 6 sparse Matrix of class "dfm"
#>        features
#> docs    document lots stopwords . , :
#>   text1        1    1         1 1 0 0
#>   text2        0    1         1 1 2 1
toks <- tokens(c("this contains lots of stopwords",
                 "no if, and, or but about it: lots"),
               remove_punct = TRUE)
tmpfcm <- fcm(toks)
tmpfcm
#> Feature co-occurrence matrix of: 12 by 12 features.
#> 12 x 12 sparse Matrix of class "fcm"
#>            features
#> features    this contains lots of stopwords no if and or but about it
#>   this         0        1    1  1         1  0  0   0  0   0     0  0
#>   contains     0        0    1  1         1  0  0   0  0   0     0  0
#>   lots         0        0    0  1         1  1  1   1  1   1     1  1
#>   of           0        0    0  0         1  0  0   0  0   0     0  0
#>   stopwords    0        0    0  0         0  0  0   0  0   0     0  0
#>   no           0        0    0  0         0  0  1   1  1   1     1  1
#>   if           0        0    0  0         0  0  0   1  1   1     1  1
#>   and          0        0    0  0         0  0  0   0  1   1     1  1
#>   or           0        0    0  0         0  0  0   0  0   1     1  1
#>   but          0        0    0  0         0  0  0   0  0   0     1  1
#>   about        0        0    0  0         0  0  0   0  0   0     0  1
#>   it           0        0    0  0         0  0  0   0  0   0     0  0
fcm_remove(tmpfcm, stopwords("english"))
#> Feature co-occurrence matrix of: 3 by 3 features.
#> 3 x 3 sparse Matrix of class "fcm"
#>            features
#> features    contains lots stopwords
#>   contains         0    1         1
#>   lots             0    0         1
#>   stopwords        0    0         0