Add new multivariate dataset to pipeline and test in notebook #72
Conversation
# Conflicts: # src/algorithms/lstm_ad.py
…rce/MP-2018 into fix/lstmad_threshold
| from .dataset import Dataset | ||
|
|
||
| """ | ||
| TODO: |
There was a problem hiding this comment.
Maybe add an issue about this (with some more details)?
There was a problem hiding this comment.
Deprecated comment removed
| """ | ||
|
|
||
|
|
||
| def get_random(x, strength=1): |
There was a problem hiding this comment.
Maybe "add_scaled_random" or something would be a more suitable method name?
|
Nice! Do you mind adding the datasets to main.py? |
| self.mean_curve_length = mean_curve_length | ||
| self.mean_curve_amplitude = mean_curve_amplitude | ||
| self.global_noise = 0.1 # Noise added to all dimensions over the whole timeseries | ||
| self.dim2 = dim2 |
There was a problem hiding this comment.
Might be better to call it "anomaly_function". At the moment there are only 2D anomalies, another PR will solve that
| # The last two values are ignored for generation of not anomalous data | ||
|
|
||
|
|
||
| def doubled_dim2(curve_values, anomalous, interval_length): |
There was a problem hiding this comment.
In the next 4 methods you're not using interval_length. I think it'd make sense to rename it to _ to indicate that it is not used.
| if not anomalous: | ||
| return curve_values, -1, -1 | ||
| else: | ||
| # The curve in the second dimension occures a few timestamps later |
| # Add anomaly labels with slight padding (dont start with the first interval value). | ||
| # The padding is curve_length / padding_factor | ||
| if create_anomaly: | ||
| assert end > start and start >= 0, f'Invalid anomaly indizes: {start} to {end}' |
There was a problem hiding this comment.
You can simplify that to assert end > start >= 0
|
Do you mind uncommenting |
Addresses #40
Comments are missingCode might be confusing -> I'll refactor it with the purpose of reusing it for other types of MV outliers.UPDATE:
Already implemented anomalies: