Appendix. 데이터 상관관계 한 눈에 파악하기

1. Enumerate, Suplot

sns.histplot(x=df['Age'])

sns.histplot(x=df['DistanceFromHome'])

for i, col in enumerate(['Age', 'DistanceFromHome', 'JobSatisfaction', 'WorkLifeBalance', 'MonthlyIncome', 'YearsAtCompany']):
  print(i, col)
# 결과값:
0 Age
# 1 DistanceFromHome
# 2 JobSatisfaction
# 3 WorkLifeBalance
# 4 MonthlyIncome
# 5 YearsAtCompany

hist = ['Age', 'DistanceFromHome', 'JobSatisfaction', 'WorkLifeBalance', 'MonthlyIncome', 'YearsAtCompany']

plt.figure(figsize=(10, 20))
for i, col in enumerate(hist):
  axes = plt.subplot(6, 3, i+1)
  sns.histplot(x=df[col], hue=df['Gender'])

plt.tight_layout()
plt.show()

hist = ['Age', 'DistanceFromHome', 'JobSatisfaction', 'WorkLifeBalance', 'MonthlyIncome', 'YearsAtCompany']

plt.figure(figsize=(10, 20))
for i, col in enumerate(hist):
  axes = plt.subplot(6, 3, i+1)
  sns.histplot(x=df[col], hue=df['Attrition'])

plt.tight_layout()
plt.show()

2. Jointplot

seaborn.jointplot parameters

sns.jointplot(data, x, y
			  , hue=[범례]
              , marker=[main data scatter 마커표시]
              , kind=[Scatter 외, kde, reg, hist 등 다른 종류의 그래프로 변경]
              , marginal_ticks=True [Marginal Data에 y축 추가]
              , marginal_kws=[Marginal Data의 세부 디자인 조정])

marker: default '.' - Dot / 'o' - Circle / '+' - Plus / '^' - Triangle / 's' - Squre / ' ' - Star / 'D' - Diamond

sns.jointplot(df, x='Age', y='YearsAtCompany', kind='reg')

j = sns.jointplot(df, x='YearsAtCompany', y='Age'
                  , marker='+'
                  , marginal_ticks=True
                  , marginal_kws=dict(bins=30, rug=True)
                  )

j = sns.jointplot(df, x='YearsAtCompany', y='Age'
                  , marker='+'
                  , marginal_ticks=True
                  , marginal_kws=dict(bins=30, rug=True)
                  )
j.plot_joint(sns.kdeplot, color='r')

j = sns.jointplot(df, x='YearsAtCompany', y='Age'
                  , marker='+'
                  , marginal_ticks=True
                  , marginal_kws=dict(bins=30, rug=True)
                  )
j.plot_joint(sns.kdeplot, color='r')
j.plot_marginals(sns.rugplot, color='r', height=-.15, clip_on=False)