CODEWITHSUNDEEP
rdictionaryquanteda
MG Fern
MG Fern

Reputation: 119

R see list of words in dictionary

I want to see the words included in a dictionary. Here is my dictionary:

Name               Type                             Value
dict_lg            list [2] (quanteda::dictionary2) List of length 2
   NEGATIVE        character [2867]                 'à côrnes' 'à court de personnel'
   POSITIVE        list [1] (quanteda::dictionary2) List of length 1 
      VÉRITÉ* (1)) character [0]

I would like to see the words included in each list (NEGATIVE, POSITIVE). If I do:

dict_lg <- dictionary(file = "frlsd/frlsd.cat", encoding = "UTF-8")
dict_lg$NEGATIVE

it prints me the list of negative words, but if I do:

dict_lg$POSITIVE

I obtain:

Dictionary object with 1 key entry.
- [VÉRITÉ* (1))]:

or if I do

dict_lg[["POSITIVE"]][["VÉRITÉ* (1))"]]

I obtain

character(0)

How can I see the list of positive words? The original dictionary file is this one: https://www.poltext.org/fr/donnees-et-analyses/lexicoder

Upvotes: 0

Views: 106

Answers (2)

Ken Benoit
Ken Benoit

Reputation: 14902

The problem here lies with the file you referenced at https://www.poltext.org/fr/donnees-et-analyses/lexicoder. For the value "VÉRITÉ" under the key "POSITIVE", it has an extra ")". Eliminate that, and the dictionary will behave properly.

enter image description here

I eliminated the extra ")" and then loaded in the edited file, and it works fine.

library("quanteda")
#> Package version: 3.3.1
#> Unicode version: 14.0
#> ICU version: 71.1
#> Parallel computing: 10 of 10 threads used.
#> See https://quanteda.io for tutorials and examples.
dict <- dictionary(file = "~/Downloads/frlsd_edited.cat")

print(dict, max_nval = 6)
#> Dictionary object with 2 key entries.
#> - [NEGATIVE]:
#>   - à côrnes, à court de personnel , à l'étroit, à peine*, abais*, abandon* [ ... and 2,861 more ]
#> - [POSITIVE]:
#>   - à l'épreuve*, à la mode, abondamment, abondance, abondant*, abonde* [ ... and 1,278 more ]

head(dict$POSITIVE)
#> [1] "à l'épreuve*" "à la mode"    "abondamment"  "abondance"    "abondant*"   
#> [6] "abonde*"

head(dict$NEGATIVE)
#> [1] "à côrnes"              "à court de personnel " "à l'étroit"           
#> [4] "à peine*"              "abais*"                "abandon*"

Created on 2023-07-24 with reprex v2.0.2

Upvotes: 1

I_O
I_O

Reputation: 6921

You can examine the list structure of the dictionary like so:

rapply(dict_lg, f = \(i) i, how = 'list') |> str()

... which suggests that the structure was messed up (either at generation of the cat-file or upon import):

List of 2
 $ NEGATIVE:List of 1
  ..$ : chr [1:2867] "à côrnes" "à court de personnel " "à l'étroit" "à peine*" ...
 $ POSITIVE:List of 2
  ..$ VÉRITÉ* (1)):List of 1
  .. ..$ : chr(0) 
  ..$             : chr [1:1283] "à l'épreuve*" "à la mode" "abondamment" "abondance" ...

... however you can pull all terms from list item 'POSITIVE' like this:

rapply(dict_lg, f = \(i) i, how = 'list')$POSITIVE

edit to convert the dictionary into a dataframe of terms and sentiments to, e. g. filter out the terms of negative sentiment:

library(dplyr)

rapply(dict_lg, f = \(i) i, how = 'unlist', ) %>%
data.frame(term = .,
           sentiment = gsub('(POSITIVE|NEGATIVE).*', '\\1', names(.))
           ) %>%
filter(sentiment == 'NEGATIVE')

                           term sentiment
NEGATIVE1              à côrnes  NEGATIVE
NEGATIVE2 à court de personnel   NEGATIVE
NEGATIVE3            à l'étroit  NEGATIVE
NEGATIVE4              à peine*  NEGATIVE
NEGATIVE5                abais*  NEGATIVE
NEGATIVE6              abandon*  NEGATIVE
## truncated

Upvotes: 2

Related Questions