Add in_column to plot_anomalies, plot_anomalies_interactive

[Mr-Geekman]

IMPORTANT: Please do not create a Pull Request without creating an issue first.

Before submitting (must do checklist)

• Did you read the contribution guide?
• Did you update the docs? We use Numpy format for all the methods and classes.
• Did you write any new necessary tests?
• Did you update the CHANGELOG?

Type of Change

• Examples / docs / tutorials / contributors update
• Bug fix (non-breaking change which fixes an issue)
• Improvement (non-breaking change which improves an existing feature)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Proposed Changes

Look #298.

Related Issue

#298.

Closing issues

Closes #298.

[Mr-Geekman]

Script for test of plot_anomalies:

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

from etna.analysis import get_anomalies_density
from etna.analysis import plot_anomalies
from etna.datasets import TSDataset


def main():
    df = pd.read_csv("examples/data/example_dataset.csv", parse_dates=["timestamp"])
    df_exog = df.copy()
    df_exog.rename(columns={"target": "leak"}, inplace=True)
    df_exog["leak"] = np.log1p(df_exog["leak"])
    timestamp = df["timestamp"].sort_values().unique()
    timestamp_train = timestamp[10:-10]
    df = df[df["timestamp"].isin(timestamp_train)]
    ts = TSDataset(df=TSDataset.to_dataset(df), df_exog=TSDataset.to_dataset(df_exog), freq="D", known_future="all")

    # plot with target
    outliers = get_anomalies_density(ts=ts, distance_coef=1.0, in_column="target")
    plot_anomalies(ts=ts, in_column="target", anomaly_dict=outliers)
    plt.savefig("anomaly_target")

    # plot with not target
    outliers = get_anomalies_density(ts=ts, distance_coef=1.0, in_column="leak")
    plot_anomalies(ts=ts, in_column="leak", anomaly_dict=outliers)
    plt.savefig("anomaly_exog")


if __name__ == "__main__":
    main()

anomaly_target:

anomaly_exog:

[Mr-Geekman]

Utility for testing plot_anomalies_interactive:

import pathlib

import numpy as np
import pandas as pd

from etna.analysis import get_anomalies_density
from etna.analysis import plot_anomalies_interactive
from etna.datasets import TSDataset

ROOT_PATH = pathlib.Path(__file__).parent


def load_data():
    df = pd.read_csv(ROOT_PATH.joinpath("examples/data/example_dataset.csv"), parse_dates=["timestamp"])
    df_exog = df.copy()
    df_exog.rename(columns={"target": "leak"}, inplace=True)
    df_exog["leak"] = np.log1p(df_exog["leak"])
    timestamp = df["timestamp"].sort_values().unique()
    timestamp_train = timestamp[10:-10]
    df = df[df["timestamp"].isin(timestamp_train)]
    ts = TSDataset(df=TSDataset.to_dataset(df), df_exog=TSDataset.to_dataset(df_exog), freq="D", known_future="all")
    return ts


def plot_target():
    ts = load_data()
    params_bounds = {"window_size": (5, 20, 1), "distance_coef": (0.1, 3, 0.25)}
    method = get_anomalies_density
    plot_anomalies_interactive(
        ts=ts, in_column="target", segment="segment_a", method=method, params_bounds=params_bounds
    )


def plot_leak():
    ts = load_data()
    params_bounds = {"window_size": (5, 20, 1), "distance_coef": (0.1, 3, 0.25)}
    method = get_anomalies_density
    plot_anomalies_interactive(ts=ts, in_column="leak", segment="segment_a", method=method, params_bounds=params_bounds)

To use it just import plot_target and plot_leak from jupyter notebook.

[codecov-commenter]

Codecov Report

Merging #618 (5b33ca4) into master (f6de083) will not change coverage.
The diff coverage is 0.00%.

@@           Coverage Diff           @@
##           master     #618   +/-   ##
=======================================
  Coverage   84.54%   84.54%           
=======================================
  Files         118      118           
  Lines        5973     5973           
=======================================
  Hits         5050     5050           
  Misses        923      923           

Impacted Files	Coverage Δ
etna/analysis/plotters.py	22.14% <0.00%> (ø)

📣 Codecov can now indicate which changes are the most critical in Pull Requests. Learn more