05wk-1

seaborn
matplotlib
Author

최규빈

Published

October 5, 2022

seaborn(2)–scatterplot, mpl미세먼지팁(2)

강의영상

https://youtube.com/playlist?list=PLQqh36zP38-xHObhZzebBIEf97Er27O_x

import

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

sns scatterplot

- data

x1,y1 =  np.random.multivariate_normal([0,0],[[1,0],[0,1]],size=1000).T
x2,y2 =  np.random.multivariate_normal([2,2],[[1,0.7],[0.7,1]],size=1000).T
  • 이변량정규분포에서 샘플추출 (추출코드를 기억할 필요는 없음)
  • 특징: x1,y1은 무상관으로 x2,y2는 선형관계를 가지도록 추출

plt 복습

plt.plot(x1,y1,'o')
plt.plot(x2,y2,'o')

sns: array

sns.scatterplot(data=None,x=x1,y=y1)
sns.scatterplot(data=None,x=x2,y=y2)
<AxesSubplot:>

sns: wide df

sns.scatterplot(data=pd.DataFrame({'x':x1,'y':y1}),x='x',y='y')
sns.scatterplot(data=pd.DataFrame({'x':x2,'y':y2}),x='x',y='y')
#sns.scatterplot(data=None,x=x2,y=y2)
<AxesSubplot:xlabel='x', ylabel='y'>

  • 억지로 그리긴 했는데 이 경우는 wide하게 만든 df는 별로 경쟁력이 없음

sns: long df

x= np.concatenate([x1,x2])
y= np.concatenate([y1,y2])
cat = ['x1']*len(x1) + ['x2']*len(x2)
df2 = pd.DataFrame({'x':x,'y':y,'cat':cat})
df2
x y cat
0 2.023919 -0.400176 x1
1 1.229622 -1.763752 x1
2 -0.413211 2.293004 x1
3 -1.343073 0.404232 x1
4 1.062845 0.030775 x1
... ... ... ...
1995 2.226805 3.683857 x2
1996 2.768263 2.678292 x2
1997 2.525295 2.815478 x2
1998 1.750193 2.289812 x2
1999 1.153290 2.095922 x2

2000 rows × 3 columns

sns.scatterplot(data=df2,x='x',y='y',hue='cat') 
<AxesSubplot:xlabel='x', ylabel='y'>

sns을 이용하여 matplotlib 액시즈에 그림 그리기 (\(\star\))

예제1

fig,ax = plt.subplots(1,3,figsize=(12,4))
ax[0].plot([1,2,4,3],'--o')
sns.scatterplot(x=x1,y=y1,ax=ax[1])
sns.scatterplot(x=x1,y=y1,ax=ax[2])
sns.scatterplot(x=x2,y=y2,ax=ax[2])
ax[2].plot([1,2,4,3],'-r',lw=5)

예제2

import cv2
!wget https://upload.wikimedia.org/wikipedia/commons/0/08/Unequalized_Hawkes_Bay_NZ.jpg 
img = cv2.imread('Unequalized_Hawkes_Bay_NZ.jpg',0)
!rm Unequalized_Hawkes_Bay_NZ.jpg 
--2022-10-05 16:33:56--  https://upload.wikimedia.org/wikipedia/commons/0/08/Unequalized_Hawkes_Bay_NZ.jpg
Resolving upload.wikimedia.org (upload.wikimedia.org)... 103.102.166.240, 2001:df2:e500:ed1a::2:b
Connecting to upload.wikimedia.org (upload.wikimedia.org)|103.102.166.240|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 110895 (108K) [image/jpeg]
Saving to: ‘Unequalized_Hawkes_Bay_NZ.jpg’

Unequalized_Hawkes_ 100%[===================>] 108.30K   548KB/s    in 0.2s    

2022-10-05 16:33:57 (548 KB/s) - ‘Unequalized_Hawkes_Bay_NZ.jpg’ saved [110895/110895]

img2 = cv2.equalizeHist(img)
img.reshape(-1)
array([127, 145, 149, ..., 146, 145, 144], dtype=uint8)
fig,ax = plt.subplots(2,2,figsize=(10,5))
ax[0,0].imshow(img,vmin=0,vmax=255,cmap='gray')
sns.histplot(img.reshape(-1),ax=ax[0,1],bins=15,lw=0,kde=True,color='C1')
ax[0,1].set_xlim(0,255)
ax[1,0].imshow(img2,vmin=0,vmax=255,cmap='gray')
sns.histplot(img2.reshape(-1),ax=ax[1,1],bins=15,lw=0,kde=True,color='C1')
<AxesSubplot:ylabel='Count'>

- seaborn: figure-level vs axes-level 의 개념

ref: https://seaborn.pydata.org/tutorial/function_overview.html#figure-level-vs-axes-level-functions

mpl 미세먼지팁 (2)

축 간격조정

import matplotlib as mpl
fig, ax = plt.subplots()
ax.plot([(xi/30)**2 for xi in range(30)],'--o')
ax.xaxis.set_major_locator(mpl.ticker.MultipleLocator(3)) # 큰 눈금간격을 3으로
ax.xaxis.set_minor_locator(mpl.ticker.MultipleLocator(1)) # 작은 눈금간격을 1로

축 삭제

fig, ax = plt.subplots()
ax.plot([(xi/30)**2 for xi in range(30)],'--o')
ax.xaxis.set_major_locator(mpl.ticker.NullLocator()) # x축 눈금삭제
ax.yaxis.set_major_locator(mpl.ticker.NullLocator()) # y축 눈금삭제

축 범위조정

fig, ax = plt.subplots()
ax.plot([(xi/30)**2 for xi in range(30)],'--o')
ax.set_ylim(-1,2) 
ax.set_xlim(-5,35)
#plt.ylim(-1,2)
#plt.xlim(-5,35)
(-5.0, 35.0)

gcf, gca

- gcf

plt.plot([1,2,3,2])
fig = plt.gcf()

fig.suptitle('suptitle')
Text(0.5, 0.98, 'suptitle')
fig

- gca

fig

ax = fig.gca()
ax.set_title('title') 
fig