Separate out smoothing and plotting functions

- Create the function adjust_and_smooth() and call that within plot_sentiments() instead of having everything be within the plot function - Fix up both those functions: clean up the legend positioning and labels, fix the adjusted mean calculation, and clarify comments - Stop using the ALL_MODELS_LIST global variable and instead just write out the full list every time to make it clear in the doc strings which SentimentR model names are expected in each function - Change jockersrinker to jockers_rinker for readability and plot legends
afelleson · Mar 19, 2024 · 9681399 · 9681399
1 parent 8bb1a35
commit 9681399
Show file tree

Hide file tree

Showing 2 changed files with 348 additions and 118 deletions.
diff --git a/src/imppkg/run_sentimentr.R b/src/imppkg/run_sentimentr.R
@@ -34,11 +34,11 @@ get_sentimentr_values <- function(sentences_vector) {
 
   get_sentences_obj <- get_sentences(sentences_vector)
 
-  jockersrinker <- sentiment(get_sentences_obj, 
-                             polarity_dt=lexicon::hash_sentiment_jockers_rinker, 
-                             hypen="", amplifier.weight=0.8, n.before=5, n.after=2,
-                             adversative.weight=0.25, neutral.nonverb.like = FALSE, 
-                             missing_value = NULL)
+  jockers_rinker <- sentiment(get_sentences_obj, 
+                              polarity_dt=lexicon::hash_sentiment_jockers_rinker, 
+                              hypen="", amplifier.weight=0.8, n.before=5, n.after=2,
+                              adversative.weight=0.25, neutral.nonverb.like = FALSE, 
+                              missing_value = NULL)
 
   jockers <- sentiment(get_sentences_obj, 
                        polarity_dt=lexicon::hash_sentiment_jockers, 
@@ -72,10 +72,10 @@ get_sentimentr_values <- function(sentences_vector) {
 
   # loughran_mcdonald is for financial texts
   loughran_mcdonald <- sentiment(get_sentences_obj, 
-                                polarity_dt=lexicon::hash_sentiment_loughran_mcdonald, 
-                                hypen="", amplifier.weight=0.8, n.before=5, n.after=2,
-                                adversative.weight=0.25, neutral.nonverb.like = FALSE, 
-                                missing_value = 0)
+                                 polarity_dt=lexicon::hash_sentiment_loughran_mcdonald, 
+                                 hypen="", amplifier.weight=0.8, n.before=5, n.after=2,
+                                 adversative.weight=0.25, neutral.nonverb.like = FALSE, 
+                                 missing_value = 0)
 
   socal_google <- sentiment(get_sentences_obj, 
                             polarity_dt=lexicon::hash_sentiment_socal_google, 
@@ -84,8 +84,8 @@ get_sentimentr_values <- function(sentences_vector) {
                             missing_value = NULL)
 
   sentimentr_df <- data.frame('text_clean' = sentences_vector,
-                              'was_averaged' = restoreOldSentences(jockersrinker)$flag,
-                              'sentimentr_jockersrinker' = restoreOldSentences(jockersrinker)$sentiment_avg,
+                              'was_averaged' = restoreOldSentences(jockers_rinker)$flag,
+                              'sentimentr_jockers_rinker' = restoreOldSentences(jockers_rinker)$sentiment_avg,
                               'sentimentr_jockers' = restoreOldSentences(jockers)$sentiment_avg,
                               'sentimentr_huliu' = restoreOldSentences(huliu)$sentiment_avg,
                               'sentimentr_nrc' = restoreOldSentences(nrc)$sentiment_avg,