MultiDatasetTradingEnv incompatibility with gymnasium==1.0.0 and issues in selection of multi-dataset

[sachetz]

I came across 2 issues in the MultiDatasetTradingEnv class:

1. gymnasium==1.0.0 requires the env.reset() function to take 2 parameters, seed and options. The class currently only has seed, and this causes an error:
   TypeError: reset() got an unexpected keyword argument 'options'

2. The selection of the dataset in the multi-dataset process seems to be broken. Instead of selecting the least used dataset, it ends up selecting the first one every time.

potential_dataset_pathes = np.where(self.dataset_nb_uses == self.dataset_nb_uses.min())[0]
random_int = np.random.randint(potential_dataset_pathes.size)
dataset_path = self.dataset_pathes[random_int]

For example, if there is only one dataset that has the minimum uses, random_int always ends up being 0, and so only the first dataset is always selected - regardless of the actual usage.