Generate bin edges from an inverse normal distribution This distribution is best suited for z-normalized time series data Parameters ---------- n_bit : int, default 8 The number of bits to be used in generating the inverse normal distribution Returns -------
(n_bit=8)
| 244 | |
| 245 | @lru_cache() |
| 246 | def _inverse_norm(n_bit=8): # pragma: no cover |
| 247 | """ |
| 248 | Generate bin edges from an inverse normal distribution |
| 249 | |
| 250 | This distribution is best suited for z-normalized time series data |
| 251 | |
| 252 | Parameters |
| 253 | ---------- |
| 254 | n_bit : int, default 8 |
| 255 | The number of bits to be used in generating the inverse normal distribution |
| 256 | |
| 257 | Returns |
| 258 | ------- |
| 259 | out : numpy.ndarray |
| 260 | Array of bin edges that can be used for data discretization |
| 261 | """ |
| 262 | return norm.ppf(np.arange(1, (2**n_bit)) / (2**n_bit)) |
| 263 | |
| 264 | |
| 265 | def _discretize(a, bins, right=True): # pragma: no cover |