Description Usage Format Value The diffrprojects class family Examples
class for diffrproject
1 |
R6Class
creator object.
Object of diffrproject
Diffrproject consists of an set of R6 classes that are conencted by inheritance. Each class handles a different set of functionalities that are modular.
A class that has nothing to do per se with diffrprojects. It merely adds some basic features to the base R6 class (debugging, hashing, getting fields and handling warnings and messages as well as listing content). This class is imported from rtext package
[inherits from rtext::R6_rtext_extended] This class forms the foundation of all diffrpojects (dp_xxx) classes by implementing data fields for meta data, texts, data on texts, links between texts, alignment of text tokens, and data on the alignment of text tokens. Furthermore it implements methods add, delete, code, and link texts or to aggregate text data on text token level.
[inherits from dp_base] This class allows for loading and saving diffrprojects from and to Rdata files.
[inherits from dp_loadsave] This class provides methods for exporting and importing to and from RSQLite.
[inherits from dp_export] This is one of the workhorses of diffrprojects. The methods of this class allow for adding, deleting or computing alignments between text tokens (e.g. words or lines or sentences or characters or paragraphs, or some other way to split text into chunks). Furthermore it allows to also assign data to individual alignments (a connection beween two token of text from different text versions).
[inherits from dp_align] The text_data_inherit method added by this class allows to copy text data from one token of a text version to another token of another text version channeled through aligments with zero distance. Conflicting codings (a text might have multiple codings stemming from several links and from direct coding of the text) are resolved by the fact that text codings are accompanied by a hierarchy level that defaults to zero and gets decreased by one every time the coding is inherited by a token.
[inherits from dp_inherit] Just a wrapper inheriting from dp_inherit to have a less technical name at the end of the inheritance chain.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 | ## Creating a Diffrprojects Instance
# To create a diffrproject we use the diffrproject creator object -
# its simply an object with an function that knows how to create a project.
# Creating a project looks like this:
library(diffrprojects)
dp <- diffrproject$new()
# Et viola - we created a first, for now empty, project that we will
# use throughout the tutorial.
## Some Help Please
# To get a better idea about what this thing called *diffrproject* really is
# you can consult its help page which gives a broad overview over its
# capabilities:
?diffrproject
# Another way is to call the ls() method. This will present us with a
# data frame listing all fields where data is stored and all the methods
# (aka object specific functions) of our diffrprojects instance.
# Those methods and fields located in *private* are not for the user
# to mess around with while non-private (*self* aka public) data fields
# can be read by the user and public methods can be triggered by the
# user to manipulate the data or retrieve data in a specific format.
dp$ls()
# The base R class() function furthermore reveals from which classes the
# diffrproject class inherits:
class(dp)
## Adding Texts to Projects
# Our diffrproject (`dp`) has one method called `text_add()` that allows to
# add texts to the project. Basically the method can be used in three
# different flavors: adding character vectors, adding texts stored on disk,
# or by adding rtext objects (see rtext package:
# https://CRAN.R-project.org/package=rtext; rtext objects are the way
# individual texts are represented within diffrprojects).
# For each of these used cases there is one option:
# `text`, `text_file`, `rtext`; respectively.
# Below are shown examples using each of these methods:
# **adding text files**
test_file1 <- stringb:::test_file("rc_1_ch1.txt")
test_file2 <- stringb:::test_file("rc_2_ch1.txt")
dp$text_add(text_file = c(test_file1, test_file2) )
# **adding rtext objects**
test_file <- stringb:::test_file("rc_1_ch1.txt")
rt <- rtext$new( text_file = test_file)
dp$text_add(rtext = rt)
# **adding character vectors**
test_file1 <- stringb:::test_file("rc_1_ch1.txt")
test_file2 <- stringb:::test_file("rc_2_ch1.txt")
cv <- ""
cv[1] <- text_read(test_file1, NULL)
cv[2] <- text_read(test_file2, NULL)
dp$text_add(text = cv)
# In the last case make sure to put each text in one separate line.
# Functions like readLines() or text_read() read in texts such that
# each line corresponds to one element in a character vector. With e.g.
# text_read()'s tokenize parameter to NULL the text will be read in as one
# long string.
## Piping Methods
# Now is a good time to mention a feature of diffrprojects that comes in
# handy: All functions that do not explicitly extract data
# (those usually have some 'get' as part of their name) do return the
# object itself so that one can pipe together a series of method calls.
# Consider the following example where we initiate a new diffrprojects
# instance and add two texts in just one pipe:
dp <-
diffrproject$
new()$
text_add(text_version_1, name = "version1")$
text_add(text_version_2, name = "version2")
length(dp$text)
## Getting Infos About Texts
# If we want to get some general overview about the texts gathered in our
# project, we can use the text_meta_data() method to do so.
# The method has no parameters and returns a data.frame with several
# variables informing us about its source, length, encoding used for
# storage, and its name.
dp$text_meta_data()
## Showing Text
# If you want to have a look at your texts you may do so by using the
# text's own text_show methods. Per default this method only shows the
# first 500 characters, but it can be set to higher numbers as well.
dp$text$version1$text_show(length=1000)
dp$text$version2$text_show(length=1000)
## Getting And Setting Infos About the Project
# Similar to the text_meta_data() method we can access the projects
# meta data via data fields meta and options. But contrary to the
# text_meta_data() method that gathers data from all the texts within the
# project and does not allow for manipulation of the data, the data
# fields allow reading and writing.
# First let us have a look and thereafter turn off the message
# notification service:
# **getting data fields**
dp$options
# **setting data fields**
dp$options$verbose <- FALSE
# (note, ask is deprecated and only remains for compatibility
# reasons but has no function anymore)
# Now it's time to have a look at the projects meta data.
# It tells us when the project was created, which path to use for
# SQLite exports, which path to use for saving data as in RData
# format and what is the projects id. The id is a hash of a time stamp
# as well as session information which should ensure uniqueness across
# space and time.
# All these values can be manipulated by the user to her liking.
dp$meta
dp$meta$file_path = "./diffrproject.RData"
## Deleting Texts
# Of cause we can not only add texts but delete them from the project as
# well. For this purpose there is the text_delete() method.
# Let's just add two texts and delete one by providing its index number and
# the second by providing its name to the text_delete() method.
dp$text_add(text = "nonesense", "n1")
dp$text_add(text = "nonesense", "n2")
dp$text_delete(3)
dp$text_delete("n2")
length(dp$text)
names(dp$text)
## Defining Relationships Between Texts: Linking
# The purpose of diffrprojects is to enable data collection on the
# difference of texts. Having filled a project with various texts,
# there are endless possibilities to form pairs of text for comparison
# and change measurement - where endless actually is equal to: $n^2-n$.
# Linking can be done via the text_link method which accepts either
# index numbers or text names for its from and to arguments
# (a third argument delete will delete a specified link if set to TRUE).
dp$text_link(from = 1, to = 2)
dp$text_link(from = 1, to = 2, delete = TRUE)
# If no arguments are specified, text_link will link the first text to
# the second, the third to the fourth, the fourth to the fifths and so on.
dp$text_link()
# To get an idea of what links are currently specified, we can
# directly access the link data field or/and ask R to transform the
# list found there into a data.frame.
dp$link
dp$link %>% as.data.frame()
## Aligning Texts and Measuring Change
# At the heart of each diffrproject lies the text_align method.
# This method compares two texts and tries to align parts
# of one text with parts of the other text. The first two
# arguments (`t1` and `t2`) are for specifying which pair
# of texts to compare - if left as-is, all text pairs that
# are specified within the link data field will be aligned.
# Text parts are arbitrary character spans defined by the
# `tokenizer` argument. This argument expects a function splitting
# text into a token data.frame. If the tokenizer argument
# is left as-is, it will default to text_tokenize_lines function
# from the stringb package.
# Text tokens can be pre-processed before alignment. The `clean`
# argument allows to hand over a function tranforming a charactr
# vector of text tokens into their clean counterparts.
# The `ignore` arguments expects a function that is able to
# transform a character vector of tokens into a logical vector
# of same length, indicating which tokens to ignore throughout
# the alignment process and which to consider.
# The next argument - `distance` - specifies which distance
# metrics to use to calculate distances between strings.
# Since the text_align method basically is a wrapper around
# diff_align you can get more information via `?diff_align`
# and since again diff_align is a wrapper around stringdist
# from the stringdist package `?stringdist::stringdist` and
# also ``?stringdist::`stringdist-metrics` `` will provide
# further insights about possible metrics and how to use the
# rest of the arguments to text_align (these are passed through
# to stringdist).
# Let's have an example using the Levenshtein distance to
# calculate distances between tokens (lines per default).
# Furthermore we allow the distance between two aligned tokens
# to be as large as 15. Tokens which do not find a partner
# below that distance are considered to have been deleted
# or respectively inserted. Tokens which find a partner with
# a non-zero distance which is not above the threshhold are
# considered changes - transformations of one token into the other.
# The following shows the resulting list of alignment data.frames.
dp$text_align(distance = "lv", maxDist = 15)
dp$alignment
# To measure the change between those two texts we can e.g. aggregate
# the distances by change type:
sum_up_changes <- function(x){
x %>%
dplyr::group_by(type) %>%
dplyr::summarise(sum_of_change = sum(distance))
}
lapply( dp$alignment, sum_up_changes)
## Coding Texts
# Now let us put some data into our diffrproject.
# The most basic method to do so is simply called text_code.
# Text_code takes up
# to five arguments (the first three are mandatory), where one
# specifies the text to be coded (`text`, either by index
# number or by name), how the variable to store the information
# is called (`x`), and the index number or a vector of those
# indicating which characters of the text should be coded.
# The last two parameters are optional and specify which value
# the variable should hold (`val`) and at which hierarchy
# level the coding is placed (`hl`, higher or equal hierarchy
# levels will overwrite existing codings of lower hierarchy
# level for the same text, character span, and variable).
dp$text_code(text = 1, x = "start", i=1:5, val = TRUE, hl = 0)
dp$text_code(text = "version2", x = "start", i=1:5, val = TRUE, hl = 0)
# The text_code method is quite verbose and in most cases more suited
# to be accessed by a machine or algorithm than by a human.
# Therefore, there are three other methods to code text:
# text_code_regex, text_code_alignment_token,
# text_code_alignment_token_regex.
# The text_code_regex method allows to search for text patterns and
# code a whole pattern instead of assigning codes character by
# character - the `i` argument of text_code gets replaced by a
# `pattern` argument. The in addition further arguments can be
# passed to the pattern search functions via `...` - see e.g.
# `?grep` for possible further arguments and
# https://stat.ethz.ch/R-manual/R-devel/library/base/html/regex.html for a
# description of regular expressions in R.
# In this example we are searching for the word *"it"* in text 1 and code
# each instance.
dp$text_code_regex(text = 1, x = "it", pattern = "\\bit\\b", ignore.case=TRUE)
# Another variant of coding text is by using alignment tokens.
# Having alignment data availible, this allows for selecting:
# link, alignment and text while the other arguments from above stay the
# same.
# having a look at alignment number 4
dp$alignment[[1]][4,]
# coding text connected by alignment number 4
dp$text_code_alignment_token(
link = 1,
alignment_i = 4,
text1 = TRUE,
text2 = TRUE,
x = "token_coding",
val = 4,
hl = 0
)
## Getting Text Codings
# The most basic way to get text data is to use the text_data method.
# This method will go through all or only selected texts, gather all
# the data stored there and put it into a neat data.frame where name
# identifies the text from which the data comes per name, char informs
# us about the character that was coded, and i refers to the characters
# position within the text. All other variables hold the data we added
# during the examples above.
dp$text_data(text = 1) %>% head()
## Aggregating Text Codings
# The usage of text_data has its merits but often one is more
# interested in text data aggregated to a specific level.
# The following three aggregation functions offer a solution
# to this problem: tokenize_text_data_lines, tokenize_text_data_words,
# and tokenize_text_data_regex. These three methods make use
# of the similiary named methods provided by the rtext package.
# One important thing to keep in mind is that using these methods
# implies aggregating several data values on character level
# into one data value at token level. Therefore there has
# to be some aggregation function to be involved. The default
# is to use the value that occurs most often on character
# level, if more than one distinct values occur more than
# once the first is choosen.
# The aggregation function can be changed to whatever function the
# user seems appropriate by passing it to `aggregate_function`
# - as long as it
# reduces a vector of values into a vector with only one value.
# The `join` argument allows to decide how text and data are
# joined
# into the resulting data.frame - left: all token, right: all data, full:
# token with or without data and data with or without token.
dp$tokenize_text_data_lines(
text = 1,
join = "right",
aggregate_function =
function(x){
paste(x[1:3], collapse = ",")
}
)
## Text Coding Inheritence
# Having aligned two texts via token pairs another functionality of
# diffrprojects becomes availible: text coding inheritance via no-change
# tokens. This means that text codings can get copied to those tokens they
# are aligned with, given that they are considered the same - i.e. the
# distance equals zero and the change type therefore is no-change.
# To show this feature we use the text_inherit method and we will
# start with a fresh example. A new project with two texts. The first text
# gets some codings, then they are aligned, and in a last step codings are
# transfered from one text to the other via the text_data_inherit method.
dp <-
diffrproject$new()$
text_add(text_version_1)$
text_add(text_version_2)$
text_code_regex(
text = 1,
x = "test1",
pattern = "This part.*?change",
val = "inherited"
)$
text_code_regex(
text = 1,
x = "test2",
pattern = "This part.*?change",
val = "inherited"
)
dp$tokenize_text_data_lines(1)
dp$
text_link()$
text_align()$
text_data_inherit(
link = 1,
direction = "forward"
)
dp$tokenize_text_data_lines(2)
## Saving and Loading Projects
# Diffrprojects also allow for storing and loading project to and
# from disk.
# note, uncomment code lines to run
# save to file
# dp$save(file = "dp_save.RData")
# remove object
# rm(dp)
# create new object and load saved data into new object
# dp <- diffrproject$new()
# dp$load("dp_save.RData")
# dp$tokenize_text_data_lines(2)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.