get_anomalies_confidence_interval should expect Pipeline

[iKintosh]

🚀 Feature Request

get_anomalies_confidence_interval should work with pipeline instead of model.

Motivation

This should expand possible option to find outliers.

Proposal

1. In get_anomalies_confidence_interval:

• Change parameter model to pipeline: Pipeline
• Set pipeline.quantiles equals to (lower_p, upper_p)

2. In PredictionIntervalOutliersTransform:

• Change parameter model to pipeline: Pipeline in constructor
• Change parameter model to pipeline: Pipeline in detect_outliers method

Test cases

No response

Alternatives

No response

Additional context

No response

Checklist

• [ ] I discussed this issue with ETNA Team

[alex-hse-repository]

Waiting for in_sample predictions in Pipeline