do_predict: Make predictions

In tonyelhabr/xengagement: Quantifying Twitter Engagement to xGPhilosophy

Make predictions

Description

Make predictions for new tweets, appending to existing predictions and SHAP values.

Usage

do_predict(
  tweets_transformed,
  stem = get_valid_stems(),
  overwrite = TRUE,
  fit = NULL,
  path_fit = NULL,
  ...,
  .overwrite = list(preds = NULL, shap = NULL),
  .path = list(preds = NULL, shap = NULL)
)

Arguments

tweets_transformed	Tweets retrieved from retrieve_tweets() and passed into transform_tweeets()
stem	Either "favorite" or "retweet"
overwrite	Whether to overwrite existing fit, predictions, and SHAP values.
fit	xgboost fit
path_fit	If fit is not provided, then the fit is imported from path_fit.
...	Unused
.overwrite	Specific booleans for overwriting specific outputs saved to file. Default is to use same value as overwrite. Only use this if you know what you're doing.
.path	Specific paths to export the predictions and SHAP values to. If left as NULL, they are saved as "preds_{stem}.rds" and "shap_{stem}.rds" in the directory at getOption("xengagement.dir_data"). As with .overwrite, only use this if you know what you're doing.

tonyelhabr/xengagement documentation built on June 22, 2022, 5 a.m.