A box-and-whisker plot, or boxplot, summarizes a numeric distribution with its first quartile, median, third quartile, whiskers, and separate outlier points. It helps inspect data quality and compare numeric features during exploratory analysis. section 2.4.2
Read a boxplot by checking:
- The box boundaries, which mark Q1 and Q3, with the median inside the box. section 2.4.2
- The whiskers, which extend to the farthest points allowed by the stated IQR rule. section 2.4.2
- Separate points beyond the whiskers, which reveal the presence and number of potential outliers. section 2.4.2
- Standardized values when comparing features measured in different units. section 2.4.2
Auto MPG numeric features
Standardize mpg, displacement, horsepower, weight, and acceleration, then draw boxplots together. The resulting chart supports inspection of unusually distributed values and possible collection or aggregation errors. section 2.4.2
standardized = ((data[numeric_feats] - data[numeric_feats].mean())
/ data[numeric_feats].std())
standardized.boxplot(column=numeric_feats, figsize= (12, 4))
A pandas listing shows how to standardize selected numeric features and render their boxplots together. section 2.4.2
