README.md

In bmmedinac/gleasonCalc: Gleason compute

gleasonCalc

This packages compute the v value (i.e. systemic value) from a (sub)system in a systemic functional network

Overview

gleasonCalc is a package to evaluate systemic functional valeur from a system network (see Figuredo & Figueredo, 2019). It gets agnation, delicacy and valeur magnitudes in data frames in order to visualize, correlate and sort systemic options in a gleason scale.

Installation

gleasonCalc is still under development. You can install development version of gleasonCalc from GitHub:

# Install development version from GitHub:
install.packages("devtools")
devtools::install_github("bmmedinac/gleasonCalc")

Usage

As a systemic-functional-linguistics-oriented package, gleasonCalc requires the transformation of a systemic network into a special format in a spreedsheet. Take for example the next English MOOD system network:

In a spreedsheet, in a column named 'features', each oposition and delicacy level must be expressed. Opositions are expressed by '/' (e.g. indicative/imperative). Delicacy levels are expressed by ':' (e.g. indicative:interrogative:wh-). So, further delicacy levels must be expressed with its all lesser delicacy levels:

#   features
1   imperative/indicative
2   indicative:declarative/indicative:interrogative
3   indicative:interrogative:yesno/indicative:interrogative:wh

At the moment, gleasonCalc just can evaluate valeur from system networks without co-selections. Any try to evaluate this kind of network will return error. Meanwhile, a co-selective system network in spreedsheet would look like this with '&' marking co-selection:

#   features
1   imperative/indicative & positive/negative
2   indicative:declarative/indicative:interrogative & positive/negative
3   indicative:interrogative:yesno/indicative:interrogative:wh & positive/negative

getFeatures,agnation and valeurCalc2

gleasonCalc have two built-in systems: mood_spanish (see Quiroz, 2013) and mental_german (see Figueredo & Figueredo, 2019). First one is a system without co-selection. Second one is a system with co-selection.

Actually, gleasonCalc has three main functions: getFeatures, agnation and valeurCalc2. It's necessary to export it from library(gleasonCalc) previously to use it: 1. getFeatures returns all the features from the input system network. At the moment, can't return co-selective features. Example of getFeatures is below:

library(gleasonCalc)
getFeatures(mood_spanish)
        features
1          menor
2          major
4     indicative
6     imperative
9    informative
12 interrogative
15        jusive
18     hortative
21      optative
25         polar
29     elemental
33   declarative
37   exclamative

1. agnation returns agnation magnitude of every option in a system network. At the moment, can't return agnation of co-selective features. Example of agnation is below:

#library(gleasonCalc)
agnation(mood_spanish)
# A tibble: 13 x 2
   features            agnation
   <fct>                <dbl>
 1 menor                    2
 2 major                    2
 3 indicative               2
 4 imperative               2
 5 informative              2
 6 interrogative            2
 7 jusive                   9
 8 hortative                9
 9 optative                 9
10 polar                    2
11 elemental                2
12 declarative              2
13 exclamative              2

1. valeurCalc2 returns a data frame with agnation, delicacy and valeur of every option in a system network. Example of valeurCalc2 is below:

#library(gleasonCalc)
valeurCalc2(mood_spanish)
# A tibble: 13 x 4
   features      agnation delicacy valeur
   <fct>            <dbl>    <dbl>  <dbl>
 1 menor                2        1      2
 2 major                2        1      2
 3 indicative           2        4      8
 4 imperative           2        4      8
 5 informative          2        9     18
 6 interrogative        2        9     18
 7 jusive               9        9     81
 8 hortative            9        9     81
 9 optative             9        9     81
10 polar                2       16     32
11 elemental            2       16     32
12 declarative          2       16     32
13 exclamative          2       16     32

valeurCalc2 can also sort valeur with sorted = TRUE argument:

valeurCalc2(mood_spanish, sorted = TRUE)
# A tibble: 13 x 4
   features      agnation delicacy valeur
   <fct>            <dbl>    <dbl>  <dbl>
 1 menor                2        1      2
 2 major                2        1      2
 3 indicative           2        4      8
 4 imperative           2        4      8
 5 informative          2        9     18
 6 interrogative        2        9     18
 7 polar                2       16     32
 8 elemental            2       16     32
 9 declarative          2       16     32
10 exclamative          2       16     32
11 jusive               9        9     81
12 hortative            9        9     81
13 optative             9        9     81

langspace

GleasonCalc2package also include functions to plot interactive 3D graph. One of this functions is langspace. This function plots a language space (see Figueredo & Figuredo, 2019) from three systems. As example, GleasonCalc2 have a built-in data frame called language_space. In order to plot a language space, dataframe must have at least three systems (in this case, TRANSITIVITY trans, MOOD mood, THEME theme.

head(language_space)
# A tibble: 6 x 7
  ID    genre        clause_ID clause                                          trans  mood theme
  <chr> <chr>            <dbl> <chr>                                           <dbl> <dbl> <dbl>
1 DVM 1 Procedimien…         1 Use a Balloon                                     232    25     9
2 DVM 1 Procedimien…         2 to Amplify Sound                                  232     9     0
3 DVM 1 Procedimien…         3 Small sounds can still make a big noise           232   297    65
4 DVM 1 Procedimien…         4 when you use a good sound conductor.              232    17    67
5 DVM 1 Procedimien…         5 Experiment with a balloon, compressed air and …   232    25     9
6 DVM 1 Procedimien…         6 to find out                                       232     9     0

From this dataframe, we can plot with langspace()function. langspace() can take seven arguments, but just four are mandatory:

langspace(df, x, y, z, grouping, instance = FALSE)

Arguments
df          The dataframe with a data.
x           First variable of the plot (e.g. TYPE OF PROCESS)
y           Second variable of the plot (e.g. MOOD)
z           Third variable of the plot (e.g. THEME)
grouping    Empty by default. Create color groups from levels of a factor (e.g. genres, fields, tenors, etc)
instance    FALSE by default. If provided, hover shows the textual instances of x, y and z  

See the next use of langspace:

langspace(language_space,                               #example dataframe
          x = language_space$trans,                     #first mandatory system
          y = language_space$mood,                      #second mandatory system
          z = language_space$theme,                     #third mandatory system
          grouping = language_space$genre,              #grouping by genre
          instance = language_space$clause,             #show instance saved in clause column
          framing = language_space$clause_ID)           #unfolding of text saved in clause_ID colum

From the example above, langspace plots the next interactive 3D graph:

%2016.27.49.png)

Actually, x-, y- and z-axis are statically defined as TYPE OF PROCESS, MOOD and THEME. Dinamyc name axis are under develope.

Contributing

Further readings

• Figueredo, G., & Figueredo, G. P. (2019). A systemic dynamics model of text production. Journal of Quantitative Linguistics. https://doi.org/10.1080/09296174.2019.1567301
• Quiroz, B. (2013). The interpersonal and experiential grammar of chilean spanish: Towards a principled systemic-functional description based on axial argumentation [PhD Thesis]. University of Sydney.

bmmedinac/gleasonCalc documentation built on Dec. 31, 2020, 8:56 p.m.