plot_group_term_network: Function to view top terms associated with a categorical...

In jpcompartir/JPackage: What the Package Does (One Line, Title Case)

Function to view top terms associated with a categorical variable.

Description

Counts the top terms that appear in posts with a grouping variable. Individual instances of grouping variable (e.g. Nike, Adidas) who have common words in their top n_terms will be placed by graph-drawing algorithm closer to those they have none in common with. Terms common to all or multiple grouping variables will be drawn towards the middle of the network. Terms that are uniquely associated with a specific grouping variable will be draw outwards towards the periphery of the network.

Usage

plot_group_term_network(
  data,
  group_var = brand,
  text_var = message,
  n_terms = 20,
  with_ties = FALSE
)

Arguments

data	Your data frame - should have text variable and a grouping variable coded in.
group_var	The variable you wish to compare, should be a categorical or factor e.g. brand, entity, sentiment, country
text_var	The text variable
n_terms	Number of terms to select for each
with_ties	Whether to allow for > n_terms if terms have equal frequency in group_var's count

Value

A network visualisation of terms associated with a grouping variable

Examples

## Not run: 
Example 1:
plot_group_term_network(data, group_var = brand, text_var = text, n_terms = 15)

Example 2:
plot_group_term_network(my_data_frame, group_var = sentiment, text_var = message, n_terms = 15)

## End(Not run)

jpcompartir/JPackage documentation built on March 20, 2023, 4 a.m.