booster | 🚀Optimizer for mobile applications | Computer Vision library

Problem #1: cooldown.has(message.author.id) && member.roles.cache.has("830503345251680298") does not return a number

Problem #2: it looks like the result returns a number in milliseconds so you need to work out the remaining time in milliseconds using the correct calculations. Instead of Date.now() you can use performance.now().

How it works:

Have you enabled members intent in the code e.g.

Yes If you are looking for a simple fix, you lower the value of scale_pos_weight. This will lower false positive rate even though your dataset is balanced.

For a more robust fix, you will need to run hyperparamter tuning search. Especially you should try different values of : scale_pos_weight, alpha, lambda, gamma and min_child_weight. Since they are the ones with the most impact on how conservative the model is going to be.

The problem is in this line

Weird. Here's what I'd try, in order:

1. Before you start training, use get_params/get_xgb_params() on your XGBRegressor model to make sure it actually used the random_state parameter you passed in. Ditto, look at the verbose log to make sure training used it.
2. Look at the target variable y. Is its distribution very weird, non-continuous? Please show us a plot or histogram? or at least some summary statistics (min, max, mean, median, sd, 1st and 3rd quartiles)? Is the un-stratified split affecting your training? (show the descriptive statistics before and after split, also on the eval set, these three sets shouldn't differ wildly). Is it easier to try to model log(y), sqrt(y), exp(y) or somesuch? Can you debug which rows are contributing to the CV error?
   • maybe eval_metric='rmse' and default objective reg:squarederror are very unsuitable to your target variable y. We should be able to tell from the plot of y, but try other [eval_metrics and objectives] (https://xgboost.readthedocs.io/en/latest/parameter.html#learning-task-parameters) like eval_metric=='logloss', 'mae', 'rmsle', etc. See that doc for full list of objective, eval_metric.
3. Also, for full determinism, set nthread parameter to 1 (single-core). The default is nthread==-1 (use all cores). Then rerun runs 1 and 2 and update the results in your question.
4. Fail all that, can you make a reproducible example (MCVE)? Either tell us where your dataset is publicly-available (source URL, not cloud link please), or make a reproducible example using any publicly-available dataset?

The problem is that message.author returns a User and only GuildMembers have roles. You should use message.member instead that returns the author of the message as a GuildMember, so you can use its roles property.

You could also simplify that if else statement using an OR operator. If there is no member mentioned the value of winner will be the author of the message:

Administrator\workspace\xgboost-win64_release_1.2.0

Support for JSON was introduced in XGBoost 1.3.

See simple function at the beginning of following code, this function filters (removes) error entries. Usage of this function at the end of code.

Try it online!

I see a few issues with your code here.

First, you're switching between msg and message constantly. I'm going to assume that those aren't two seperate objects you're trying to access, so you should only be using one of those.

Second, you're not searching for roles correctly. Since discord.js v12, you have to use cache when requesting data from a collection (see here for more info on what changed in v12). Also, when searching for roles, you can't use ("name", "") as far as I know. In your case, you would need to use msg.member.roles.cache.has(allowedRole.id) and msg.guild.roles.cache.find(r => r.name === "・booster")

Third, you were checking the roles for the message author, not the winner. I assume you want to check if the winner has those certain roles, not the user who chose said winner.

Finally, you had a ton of issues with leaving parentheses open. That would also cause a ton of errors.

I cleaned up your code quite a bit. This should work a lot better for you