Generating 1D scatter plots with outliers

This examle will plot the raw data points in a 1D scatter plot. Here, we attempt to highlight outliers by calculating the interquartile range.

Possible outliers for "s1":

|     |     age |    sex |     bmi |      bp |   s1 |    s2 |       s3 |      s4 |    s5 |      s6 |
|-----|---------|--------|---------|---------|------|-------|----------|---------|-------|---------|
| 123 |  0.0054 |  0.051 |  0.035  | -0.0011 | 0.15 | 0.2   | -0.062   |  0.19   | 0.016 |  0.073  |
| 161 | -0.045  |  0.051 |  0.064  |  0.07   | 0.13 | 0.13  | -0.04    |  0.11   | 0.076 |  0.086  |
| 202 |  0.082  |  0.051 |  0.0013 |  0.036  | 0.13 | 0.091 |  0.019   |  0.034  | 0.084 | -0.03   |
| 230 | -0.038  |  0.051 |  0.071  | -0.057  | 0.15 | 0.16  |  0.00078 |  0.072  | 0.05  |  0.069  |
| 248 | -0.042  | -0.045 |  0.048  |  0.06   | 0.13 | 0.13  | -0.025   |  0.11   | 0.064 |  0.04   |
| 276 |  0.013  | -0.045 |  0.026  |  0.063  | 0.13 | 0.092 |  0.063   | -0.0026 | 0.058 | -0.022  |
| 287 |  0.045  | -0.045 | -0.0062 | -0.016  | 0.13 | 0.13  |  0.019   |  0.034  | 0.032 | -0.0052 |
| 346 |  0.0054 |  0.051 |  0.018  |  0.032  | 0.13 | 0.13  | -0.021   |  0.071  | 0.063 |  0.015  |

Possible outliers for "s2":

|     |     age |    sex |     bmi |      bp |   s1 |   s2 |       s3 |    s4 |    s5 |      s6 |
|-----|---------|--------|---------|---------|------|------|----------|-------|-------|---------|
| 123 |  0.0054 |  0.051 |  0.035  | -0.0011 | 0.15 | 0.2  | -0.062   | 0.19  | 0.016 |  0.073  |
| 161 | -0.045  |  0.051 |  0.064  |  0.07   | 0.13 | 0.13 | -0.04    | 0.11  | 0.076 |  0.086  |
| 230 | -0.038  |  0.051 |  0.071  | -0.057  | 0.15 | 0.16 |  0.00078 | 0.072 | 0.05  |  0.069  |
| 248 | -0.042  | -0.045 |  0.048  |  0.06   | 0.13 | 0.13 | -0.025   | 0.11  | 0.064 |  0.04   |
| 287 |  0.045  | -0.045 | -0.0062 | -0.016  | 0.13 | 0.13 |  0.019   | 0.034 | 0.032 | -0.0052 |
| 346 |  0.0054 |  0.051 |  0.018  |  0.032  | 0.13 | 0.13 | -0.021   | 0.071 | 0.063 |  0.015  |
| 376 | -0.0019 | -0.045 |  0.068  | -0.0057 | 0.12 | 0.13 | -0.025   | 0.087 | 0.046 | -0.0011 |

Possible outliers for "s3":

|     |    age |    sex |     bmi |     bp |      s1 |      s2 |   s3 |     s4 |       s5 |      s6 |
|-----|--------|--------|---------|--------|---------|---------|------|--------|----------|---------|
|  35 |  0.049 |  0.051 | -0.031  | -0.049 |  0.049  | -0.0041 | 0.13 | -0.054 |  0.021   |  0.02   |
|  58 |  0.042 | -0.045 | -0.064  |  0.036 |  0.012  | -0.058  | 0.18 | -0.076 | -0.00061 | -0.051  |
| 260 |  0.042 | -0.045 | -0.0084 | -0.057 |  0.0081 | -0.031  | 0.15 | -0.076 | -0.08    | -0.018  |
| 261 |  0.049 | -0.045 | -0.042  |  0.1   |  0.036  | -0.026  | 0.18 | -0.076 | -0.013   |  0.015  |
| 269 |  0.009 | -0.045 | -0.032  | -0.026 |  0.042  | -0.01   | 0.16 | -0.076 | -0.012   | -0.038  |
| 286 | -0.038 | -0.045 | -0.055  | -0.078 | -0.033  | -0.086  | 0.14 | -0.076 | -0.019   | -0.0052 |
| 441 | -0.045 | -0.045 | -0.073  | -0.081 |  0.084  |  0.028  | 0.17 | -0.039 | -0.0042  |  0.0031 |

Possible outliers for "s4":

|     |     age |    sex |   bmi |      bp |    s1 |     s2 |     s3 |   s4 |    s5 |    s6 |
|-----|---------|--------|-------|---------|-------|--------|--------|------|-------|-------|
| 123 |  0.0054 |  0.051 | 0.035 | -0.0011 | 0.15  |  0.2   | -0.062 | 0.19 | 0.016 | 0.073 |
| 216 |  0.013  |  0.051 | 0.036 |  0.049  | 0.053 |  0.074 | -0.069 | 0.15 | 0.046 | 0.049 |
| 322 |  0.024  |  0.051 | 0.062 |  0.062  | 0.025 | -0.036 | -0.091 | 0.16 | 0.13  | 0.082 |
| 336 | -0.02   | -0.045 | 0.085 | -0.037  | 0.092 |  0.089 | -0.062 | 0.15 | 0.081 | 0.053 |

Possible outliers for "s5":

|     |     age |    sex |    bmi |     bp |    s1 |      s2 |      s3 |    s4 |   s5 |     s6 |
|-----|---------|--------|--------|--------|-------|---------|---------|-------|------|--------|
|  23 |  0.045  |  0.051 |  0.061 |  0.031 | 0.029 | -0.047  | -0.054  | 0.071 | 0.13 | 0.14   |
| 169 | -0.0019 | -0.045 | -0.027 |  0.049 | 0.059 | -0.016  | -0.047  | 0.071 | 0.13 | 0.02   |
| 322 |  0.024  |  0.051 |  0.062 |  0.062 | 0.025 | -0.036  | -0.091  | 0.16  | 0.13 | 0.082  |
| 353 | -0.053  | -0.045 | -0.056 | -0.037 | 0.089 | -0.0032 |  0.0081 | 0.034 | 0.13 | 0.0031 |

Possible outliers for "s6":

|     |     age |    sex |    bmi |      bp |     s1 |     s2 |      s3 |      s4 |     s5 |    s6 |
|-----|---------|--------|--------|---------|--------|--------|---------|---------|--------|-------|
|  23 |  0.045  |  0.051 |  0.061 |  0.031  |  0.029 | -0.047 | -0.054  |  0.071  |  0.13  |  0.14 |
|  84 |  0.0018 | -0.045 | -0.04  | -0.1    | -0.029 | -0.03  |  0.045  | -0.05   | -0.068 | -0.13 |
| 117 |  0.06   | -0.045 | -0.021 |  0.087  |  0.045 |  0.032 | -0.047  |  0.071  |  0.079 |  0.14 |
| 141 | -0.027  | -0.045 |  0.048 | -0.047  |  0.034 |  0.057 | -0.08   |  0.13   |  0.045 |  0.13 |
| 168 |  0.0018 |  0.051 |  0.06  | -0.0022 |  0.062 |  0.063 | -0.058  |  0.11   |  0.069 |  0.13 |
| 245 | -0.027  | -0.045 | -0.035 | -0.03   | -0.057 | -0.059 |  0.03   | -0.039  | -0.05  | -0.13 |
| 350 | -0.027  |  0.051 |  0.061 |  0.11   |  0.012 | -0.018 | -0.0029 | -0.0026 |  0.07  |  0.14 |
| 406 | -0.056  | -0.045 | -0.081 | -0.085  | -0.037 | -0.037 |  0.034  | -0.039  | -0.056 | -0.14 |
| 428 |  0.049  |  0.051 |  0.089 |  0.087  |  0.036 |  0.022 | -0.025  |  0.034  |  0.066 |  0.13 |

from matplotlib import pyplot as plt
import pandas as pd
from sklearn.datasets import load_diabetes
from tabulate import tabulate
from psynlig import generate_1d_scatter
plt.style.use('seaborn-talk')


data_set = load_diabetes()
data = pd.DataFrame(data_set['data'], columns=data_set['feature_names'])

variables = ['s1', 's2', 's3', 's4', 's5', 's6']

kwargs = {
    'scatter': {
        'marker': 'o',
        's': 200,
        'alpha': 0.7,
    },
    'scatter-outlier': {
        's': 100,
        'marker': 'o',
        'label': 'Outliers',
    },
    'figure': {'figsize': (12, 6)},
}

_, _, outliers = generate_1d_scatter(
    data,
    variables,
    show_legend=True,
    outliers=True,
    **kwargs,
)
for var, out in outliers.items():
    print('\nPossible outliers for "{}":\n'.format(var))
    data_out = data.iloc[out, :]
    print(
        tabulate(data_out, tablefmt='github', headers='keys', floatfmt='.2g')
    )

plt.show()