############################ Analyse mit fastText ######################

# Korpusvorbereitung (vgl. Datei "05a Korpusvorbereitung-Fuer-fastText-Calderon.R": 
# - alle 21 identifizierten Tragödien
# - alle 21 identifiezierten Komödien
# mit Leerzeichen, Zeichensetzung, Eigennamen, Stopwords entfernt
# alles in Kleinschreibung
# die 21 Tragödien bzw. Komödien jeweils in einer einzigen Datei 
# zusammengefasst

# das Paket fastText aufrufen
library(fastText)

# Das Arbeitsverzeichnis auf den "fastText"-Ordner setzen

# Das Verzeichnis für den output in einem Vektor abspeichern
output <- getwd()

# 'skipgram' Funktion berechnet n-Grams
list_params = list(command = 'skipgram', 
                   lr = 0.1,
                   dim = 50,
                   input = "Tragedies.txt",
                   output = file.path(output, 'word_vectorsT'),
                   verbose = 2, 
                   thread = 1,
                   minn = 2, 
                   maxn = 2)

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'skipgram_logsT.txt'),
                         MilliSecs = 5)


# 'nn' Funktion gibt die Nearest Neighbors für ein bestimmtes Wort
# wieder, basierend auf dem oben erzeugten Input
# hier für den Begriff 'poder'
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'poder')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'poderT-nearest.txt'))

# amor
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'amor')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'amorT-nearest.txt'))

# honor
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'honor')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'honorT-nearest.txt'))


# Dasselbe für die Komödien
# 'skipgram' Funktion berechnet n-Grams
list_params = list(command = 'skipgram', 
                   lr = 0.1,
                   dim = 50,
                   input = "Comedies.txt",
                   output = file.path(output, 'word_vectorsC'),
                   verbose = 2, 
                   thread = 1,
                   minn = 2, 
                   maxn = 2)

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'skipgram_logsC.txt'),
                         MilliSecs = 5)

# 'nn' Funktion gibt die Nearest Neighbors für ein bestimmtes Wort
# wieder, basierend auf dem oben erzeugten Input
# hier für den Begriff 'poder'
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'poder')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'poderC-nearest.txt'))

# amor
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'amor')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'amorC-nearest.txt'))

# honor
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'honor')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'honorC-nearest.txt'))

# Nun der Vergleich der Nearest Neighbors für einige der identischen
# Begriffe in beiden Subkorpora
# celos
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'celos')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'celosT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'celos')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'celosC-nearest.txt'))


# dios
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'dios')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'diosT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'dios')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'diosC-nearest.txt'))




# guerra
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'guerra')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'guerraT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'guerra')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'guerraC-nearest.txt'))




# hado
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'hado')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'hadoT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'hado')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'hadoC-nearest.txt'))




# hado
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'hado')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'hadoT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'hado')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'hadoC-nearest.txt'))




# justicia
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'justicia')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'justiciaT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'justicia')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'justiciaC-nearest.txt'))




# muera
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'muera')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'mueraT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'muera')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'mueraC-nearest.txt'))




# villano
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'villano')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'villanoT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'villano')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'villanoC-nearest.txt'))




# amistad
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'amistad')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'amistadT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'amistad')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'amistadC-nearest.txt'))




# celos
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'celos')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'celosT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'celos')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'celosC-nearest.txt'))




# fineza
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'fineza')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'finezaT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'fineza')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'finezaC-nearest.txt'))




# gusto
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'gusto')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'gustoT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'gusto')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'gustoC-nearest.txt'))




# vida
list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsT.bin'),
                   k = 10,
                   query_word = 'vida')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'vidaT-nearest.txt'))

list_params = list(command = 'nn',
                   model = file.path(output, 'word_vectorsC.bin'),
                   k = 10,
                   query_word = 'vida')

res = fasttext_interface(list_params, 
                         path_output = file.path(output, 'vidaC-nearest.txt'))