The provided code below generates a grid of subplots (dynamic rows and 2 columns) and plots cumulative distribution function (CDF) plots for numerical variables in a DataFrame (df).

num_col = df.select_dtypes(include=np.number).columns.to_list()
col_count = len(num_col)

fig_rows = (col_count // 2 ) + 1 if col_count % 2 != 0 else col_count //2

fig, axes = plt.subplots(fig_rows, 2, figsize=(20, 15))
fig.suptitle("CDF plot of numerical variables", fontsize=20)
counter = 0
for i in range(fig_rows):
    sns.ecdfplot(ax=axes[i][0], x=df[num_col[counter]])
    counter = counter + 1
    if counter != col_count:
        sns.ecdfplot(ax=axes[i][1], x=df[num_col[counter]])
        counter = counter + 1
    else:
        pass

fig.tight_layout(pad=2.0)

GO ONE LEVEL DEEPER

How to read a CDF plot

A cumulative distribution function answers a practical question: what share of observations is less than or equal to a chosen value?

Read the chart in three moves
X

Choose a value

Start at the measurement you want to understand on the horizontal axis.

↑

Meet the curve

Move upward until you reach the empirical CDF line.

Y

Read the share

Move left to find the cumulative proportion on the vertical axis.

Keep in mind

  • Steep sections indicate many observations in a narrow range.
  • Horizontal gaps indicate ranges with few or no observations.
  • Shared axes make distribution comparisons much easier.

CONTINUE EXPLORING