using Custom Dataset

[Divya216]

I am using my custom dataset for my project - PhishingData.csv which has 11055 rows and 32 features and have converted it into dictionaries to evaluate using the below code:

dataset = [(x, x.pop("Result")) for x in df.to_dict(orient = "records")]

I would like to apply some models on my list dataset like the below from your River API reference

model = anomaly.QuantileFilter(
... anomaly.OneClassSVM(nu=0.2),
... q=0.995
... )

auc = metrics.ROCAUC()

for x, y in dataset.take(2500):
... score = model.score_one(x)
... is_anomaly = model.classify(score)
... model = model.learn_one(x)
... auc = auc.update(y, is_anomaly)

auc

In the following line if I replace datasets.CreditCard() with just my dataset it gives me the below error:

AttributeError: 'list' object has no attribute 'take'

How can i handle this list ??

[MaxHalford]

Hey hey 👋

So the datasets in the datasets module are Dataset instances. Meaning they have a .take(k) method. But your dataset is a list, so you don't have a .take(k) method. What you can do is simply slice your list:

for x, y in dataset[:2500]:
    ...

You could also create a River dataset by inheriting from the Dataset base class, but that's not as straightforward.

[Divya216]

Slicing Worked !! Thanks a ton 👍