Change occurrences of np.float_ to np.float64 as it has been removed in numpy>2.0.0

[denizberkin]

Changed all occurrences, afaik.
Should work with versions of python>=3.10 and numpy>2.0.0.

• preferred np.float64 because of high range, this can be changed in the case of needing performance without a high range limitation.
• some type-checking further on became obsolete because of the change forcing the data type to np.float64. (e.g. type(data) in [np.float64, np.float32])