sentometrics

sentometrics copied to clipboard

A possible way to deal with elisions

1

The current tokenization leaves French elisions attached to their words. This causes some sentiment words to not be identified when computing sentiment. For example, "l'abandon" is not identified as negative...

odelmarcelle

Denominator for "proportionalPol" sentiment computation

1

It always tickles me that compute_sentiment can yield values outside the [-1;1] range when using the "proportionalPol" method. ``` r library(sentometrics) sample_text id word_count LM_sample #> 1: C'est un abandon...

odelmarcelle

Danger when using `tokens` from a un-ordered corpus

5

I like tokenizing text by myself before using compute_sentiment. My usual framework is to start from a quanteda::corpus, from which I create a sento_corpus and a quanteda::tokens object. I just...

odelmarcelle

Scary cleaning V2

2

The result of a cleaning with BFG repo cleaner. The cleaning correct a large number of commits in the past.

odelmarcelle

Hi, would it be possible to implement a custom aggregation method in the `ctr_agg` function? At the moment my data is quarterly (GDP and PCE). ```r [1] "2000-02-01" "2000-05-01" "2000-08-01"...

HanjoStudy