[Visualization] seaborn

데이터1 : 비행 데이터

용량이 커서 xlsb 형식으로 저장되어 있다.

압축해제 후 엑셀 > F12 > csv 형식으로 저장하여 분석에 사용가능하다.

flights.vol1.egg

10.00MB

flights.vol2.egg

10.00MB

flights.vol3.egg

10.00MB

flights.vol4.egg

10.00MB

flights.vol5.egg

10.00MB

flights.vol6.egg

10.00MB

flights.vol7.egg

10.00MB

flights.vol8.egg

5.64MB

데이터2 : 인사 데이터

aug_train.csv

1.87MB

 

 

 

데이터 출처:

https://www.kaggle.com/datasets/usdot/flight-delays

https://www.kaggle.com/datasets/arashnic/hr-analytics-job-change-of-data-scientists?resource=download

 

 

Seaborn

https://seaborn.pydata.org/tutorial/introduction

 

In [68]:

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

 

 

catplot

 

In [69]:

df = sns.load_dataset('tips')

 

In [70]:

sns.catplot(x='day',y='total_bill', data = df)

 

Out[70]:

<seaborn.axisgrid.FacetGrid at 0x7f8b956c8d30>

 

 

 

relplot : scatter + line

 

In [71]:

penguins = sns.load_dataset('penguins'); penguins

 

Out[71]:

	species	island	bill_length_mm	bill_depth_mm	flipper_length_mm	body_mass_g	sex
0	Adelie	Torgersen	39.1	18.7	181.0	3750.0	Male
1	Adelie	Torgersen	39.5	17.4	186.0	3800.0	Female
2	Adelie	Torgersen	40.3	18.0	195.0	3250.0	Female
3	Adelie	Torgersen	NaN	NaN	NaN	NaN	NaN
4	Adelie	Torgersen	36.7	19.3	193.0	3450.0	Female
...	...	...	...	...	...	...	...
339	Gentoo	Biscoe	NaN	NaN	NaN	NaN	NaN
340	Gentoo	Biscoe	46.8	14.3	215.0	4850.0	Female
341	Gentoo	Biscoe	50.4	15.7	222.0	5750.0	Male
342	Gentoo	Biscoe	45.2	14.8	212.0	5200.0	Female
343	Gentoo	Biscoe	49.9	16.1	213.0	5400.0	Male

344 rows × 7 columns

 

In [72]:

sns.set_theme(style="whitegrid", palette="pastel")
# sns.set_theme('talk')
# sns.set_style("whitegrid")
g =sns.relplot(data = penguins, x = 'flipper_length_mm', y='bill_length_mm', col = 'island', hue='sex', style = 'species') # kind='line'
# col 인자에 범주형 칼러명을 지정하여 해당 칼럼 범주 수 만큼 그래프를 분할
g.set_xticklabels(rotation=45)
# https://stackoverflow.com/questions/32542957/control-tick-labels-in-python-seaborn-package

 

Out[72]:

<seaborn.axisgrid.FacetGrid at 0x7f8b957cc460>

 

 

 

lineplot

 

In [73]:

sns.set_theme(style = 'whitegrid')

# example
random_state = np.random.RandomState(365)
#  rns라는 object가 생기며 이를 통해 난수에 접근 가능하다
vlaues = random_state.randn(365, 4).cumsum(axis = 0)
# randn(m,n) : 평균 0, 표준편차 1의 가우시안 표준정규분포 난수를 matrix array(m,n)생성
dates = pd.date_range('2 2 2020', periods = 365, freq='D')
df = pd.DataFrame(vlaues, dates, columns= ['A','B','C','D'])
df = df.rolling(7).mean()
# rolling() : 기간 이동평균 계산

 

 

rc : runtime command
런타임 구성을 위한 명령어입니다.
matplotlib에서 다른 옵션들을 사용하게 해준다.

 

In [74]:

sns.set(rc={'figure.figsize':(5,7.49)})
sns.lineplot(data = df, palette = 'tab10', linewidth = 1 )

 

Out[74]:

<matplotlib.axes._subplots.AxesSubplot at 0x7f8b977e6df0>

 

 

In [75]:

df.fillna(1)

 

Out[75]:

	A	B	C	D
2020-02-02	1.000000	1.000000	1.000000	1.000000
2020-02-03	1.000000	1.000000	1.000000	1.000000
2020-02-04	1.000000	1.000000	1.000000	1.000000
2020-02-05	1.000000	1.000000	1.000000	1.000000
2020-02-06	1.000000	1.000000	1.000000	1.000000
...	...	...	...	...
2021-01-27	-18.482826	10.330142	-12.108625	14.878444
2021-01-28	-18.693797	10.391382	-12.020502	15.376387
2021-01-29	-18.752957	10.062616	-11.685921	15.996722
2021-01-30	-18.918042	9.957435	-11.244617	16.669990
2021-01-31	-19.239433	10.189316	-10.733219	17.307552

365 rows × 4 columns

 

In [76]:

sns.set(rc={'figure.figsize':(15,12)})
sns.lineplot(data = df, palette = 'tab10', linewidth = 3.5 )

 

Out[76]:

<matplotlib.axes._subplots.AxesSubplot at 0x7f8b977dcca0>

 

 

In [92]:

from pandas_datareader import data as pdr
import yfinance as yf
from datetime import datetime

 

 

pandas-datareader

https://pandas-datareader.readthedocs.io/en/latest/remote_data.html

pandas-datareader issue

https://github.com/pydata/pandas-datareader/issues/868

 

In [96]:

yf.pdr_override()
stock = ['GOOG','MSFT','TSLA','META']
df = pdr.get_data_yahoo(stock, start = "2019-01-01", end = "2021-04-30")

 

 

[*********************100%***********************]  4 of 4 completed

 

In [97]:

df.head(100)

 

Out[97]:

	Adj Close				Close				High		...	Low		Open				Volume
	GOOG	META	MSFT	TSLA	GOOG	META	MSFT	TSLA	GOOG	META	...	MSFT	TSLA	GOOG	META	MSFT	TSLA	GOOG	META	MSFT	TSLA
Date
2019-01-02 00:00:00-05:00	52.292500	135.679993	96.874687	20.674667	52.292500	135.679993	101.120003	20.674667	52.616001	137.509995	...	98.940002	19.920000	50.828499	128.990005	99.550003	20.406668	30652000	28146200	35329300	174879000
2019-01-03 00:00:00-05:00	50.803001	131.740005	93.310860	20.024000	50.803001	131.740005	97.400002	20.024000	52.848999	137.169998	...	97.199997	19.825333	52.049999	134.690002	100.099998	20.466667	36822000	22717900	42579100	104478000
2019-01-04 00:00:00-05:00	53.535500	137.949997	97.650703	21.179333	53.535500	137.949997	101.930000	21.179333	53.542000	138.000000	...	98.930000	20.181999	51.629501	134.009995	99.720001	20.400000	41878000	29002100	44060600	110911500
2019-01-07 00:00:00-05:00	53.419498	138.050003	97.775215	22.330667	53.419498	138.050003	102.059998	22.330667	53.700001	138.869995	...	100.980003	21.183332	53.575001	137.559998	101.639999	21.448000	39638000	20089300	35656100	113268000
2019-01-08 00:00:00-05:00	53.813999	142.529999	98.484177	22.356667	53.813999	142.529999	102.800003	22.356667	54.228001	143.139999	...	101.709999	21.801332	53.805500	139.889999	103.040001	22.797333	35298000	26263800	31514400	105127500
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2019-05-20 00:00:00-04:00	56.942501	182.720001	121.886856	13.690667	56.942501	182.720001	126.220001	13.690667	57.339851	184.229996	...	125.760002	13.016667	57.224998	181.880005	126.519997	13.520000	27066000	10352000	23706900	307893000
2019-05-21 00:00:00-04:00	57.481499	184.820007	122.543518	13.672000	57.481499	184.820007	126.900002	13.672000	57.635399	185.699997	...	126.580002	13.069333	57.424500	184.570007	127.430000	13.184000	23196000	7502800	15293300	270058500
2019-05-22 00:00:00-04:00	57.570999	185.320007	123.287079	12.848667	57.570999	185.320007	127.669998	12.848667	57.925999	186.740005	...	126.519997	12.785333	57.337502	184.729996	126.620003	13.273333	18290000	9213800	15396500	280278000
2019-05-23 00:00:00-04:00	57.038502	180.869995	121.848228	13.032667	57.038502	180.869995	126.180000	13.032667	57.298649	183.899994	...	124.739998	12.414667	57.025002	182.419998	126.199997	12.956000	23978000	12768800	23603800	398206500
2019-05-24 00:00:00-04:00	56.673500	181.059998	121.906166	12.708667	56.673500	181.059998	126.239998	12.708667	57.488251	183.630005	...	125.970001	12.583333	57.368000	182.330002	126.910004	13.322000	22240000	8807700	14123400	212049000

100 rows × 24 columns

 

In [101]:

sns.set_theme(style = 'whitegrid')
sns.set_style('darkgrid')
fig, ax = plt.subplots() 
fig.set_size_inches(12,9)

sns.lineplot(data =df['Close'], palette = 'bright', linewidth = 3.5)

 

Out[101]:

<matplotlib.axes._subplots.AxesSubplot at 0x7f8b94518970>

 

 

 

Practical example 1

 

In [124]:

cd /content/

 

 

/content

 

In [104]:

flights_df = pd.read_csv('flights.csv')

 

In [105]:

flights_df.head(10)

 

Out[105]:

	YEAR	MONTH	DAY	DAY_OF_WEEK	AIRLINE	FLIGHT_NUMBER	TAIL_NUMBER	ORIGIN_AIRPORT	DESTINATION_AIRPORT	SCHEDULED_DEPARTURE	...	ARRIVAL_TIME	ARRIVAL_DELAY	DIVERTED	CANCELLED	CANCELLATION_REASON	AIR_SYSTEM_DELAY	SECURITY_DELAY	AIRLINE_DELAY	LATE_AIRCRAFT_DELAY	WEATHER_DELAY
0	2015	1	1	4	AS	98	N407AS	ANC	SEA	5.0	...	408.0	-22.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
1	2015	1	1	4	AA	2336	N3KUAA	LAX	PBI	10.0	...	741.0	-9.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
2	2015	1	1	4	US	840	N171US	SFO	CLT	20.0	...	811.0	5.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
3	2015	1	1	4	AA	258	N3HYAA	LAX	MIA	20.0	...	756.0	-9.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
4	2015	1	1	4	AS	135	N527AS	SEA	ANC	25.0	...	259.0	-21.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
5	2015	1	1	4	DL	806	N3730B	SFO	MSP	25.0	...	610.0	8.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
6	2015	1	1	4	NK	612	N635NK	LAS	MSP	25.0	...	509.0	-17.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
7	2015	1	1	4	US	2013	N584UW	LAX	CLT	30.0	...	753.0	-10.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
8	2015	1	1	4	AA	1112	N3LAAA	SFO	DFW	30.0	...	532.0	-13.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN
9	2015	1	1	4	DL	1173	N826DN	LAS	ATL	30.0	...	656.0	-15.0	0.0	0.0	NaN	NaN	NaN	NaN	NaN	NaN

10 rows × 31 columns

 

In [108]:

flights_df.dtypes

 

Out[108]:

YEAR                     int64
MONTH                    int64
DAY                      int64
DAY_OF_WEEK              int64
AIRLINE                 object
FLIGHT_NUMBER            int64
TAIL_NUMBER             object
ORIGIN_AIRPORT          object
DESTINATION_AIRPORT     object
SCHEDULED_DEPARTURE    float64
DEPARTURE_TIME         float64
DEPARTURE_DELAY        float64
TAXI_OUT               float64
WHEELS_OFF             float64
SCHEDULED_TIME         float64
ELAPSED_TIME           float64
AIR_TIME               float64
DISTANCE               float64
WHEELS_ON              float64
TAXI_IN                float64
SCHEDULED_ARRIVAL      float64
ARRIVAL_TIME           float64
ARRIVAL_DELAY          float64
DIVERTED               float64
CANCELLED              float64
CANCELLATION_REASON     object
AIR_SYSTEM_DELAY       float64
SECURITY_DELAY         float64
AIRLINE_DELAY          float64
LATE_AIRCRAFT_DELAY    float64
WEATHER_DELAY          float64
dtype: object

 

In [111]:

flights_df['DEPARTURE_TIME'].isnull().values.any()

 

Out[111]:

True

 

In [112]:

test_flights = flights_df.fillna(1)

 

In [116]:

test_flights.columns

 

Out[116]:

Index(['YEAR', 'MONTH', 'DAY', 'DAY_OF_WEEK', 'AIRLINE', 'FLIGHT_NUMBER',
       'TAIL_NUMBER', 'ORIGIN_AIRPORT', 'DESTINATION_AIRPORT',
       'SCHEDULED_DEPARTURE', 'DEPARTURE_TIME', 'DEPARTURE_DELAY', 'TAXI_OUT',
       'WHEELS_OFF', 'SCHEDULED_TIME', 'ELAPSED_TIME', 'AIR_TIME', 'DISTANCE',
       'WHEELS_ON', 'TAXI_IN', 'SCHEDULED_ARRIVAL', 'ARRIVAL_TIME',
       'ARRIVAL_DELAY', 'DIVERTED', 'CANCELLED', 'CANCELLATION_REASON',
       'AIR_SYSTEM_DELAY', 'SECURITY_DELAY', 'AIRLINE_DELAY',
       'LATE_AIRCRAFT_DELAY', 'WEATHER_DELAY'],
      dtype='object')

 

In [123]:

sns.set_theme(style = 'whitegrid')

cmap = sns.cubehelix_palette(rot=-.3, as_cmap=True)

g = sns.relplot(data = test_flights, x="DISTANCE", y='AIR_TIME', hue = "FLIGHT_NUMBER", size = 'DEPARTURE_DELAY',palette = cmap )

g.fig.set_size_inches(15,10)
g.set(xscale = 'log', yscale = 'log')
g.ax.xaxis.grid(True, 'minor', linewidth = .25)
g.ax.yaxis.grid(True, 'minor', linewidth = .25)
g.despine(left = True, bottom = True)# axis tick mark들을 연결하는 선 제거

 

Out[123]:

<seaborn.axisgrid.FacetGrid at 0x7f8b919bafd0>

 

 

 

Practical example 2

 

 

jointplot

 

In [125]:

hr_df = pd.read_csv('aug_train.csv')

 

In [126]:

hr_df.head(10)

 

Out[126]:

	enrollee_id	city	city_development_index	gender	relevent_experience	enrolled_university	education_level	major_discipline	experience	company_size	company_type	last_new_job	training_hours	target
0	8949	city_103	0.920	Male	Has relevent experience	no_enrollment	Graduate	STEM	>20	NaN	NaN	1	36	1.0
1	29725	city_40	0.776	Male	No relevent experience	no_enrollment	Graduate	STEM	15	50-99	Pvt Ltd	>4	47	0.0
2	11561	city_21	0.624	NaN	No relevent experience	Full time course	Graduate	STEM	5	NaN	NaN	never	83	0.0
3	33241	city_115	0.789	NaN	No relevent experience	NaN	Graduate	Business Degree	<1	NaN	Pvt Ltd	never	52	1.0
4	666	city_162	0.767	Male	Has relevent experience	no_enrollment	Masters	STEM	>20	50-99	Funded Startup	4	8	0.0
5	21651	city_176	0.764	NaN	Has relevent experience	Part time course	Graduate	STEM	11	NaN	NaN	1	24	1.0
6	28806	city_160	0.920	Male	Has relevent experience	no_enrollment	High School	NaN	5	50-99	Funded Startup	1	24	0.0
7	402	city_46	0.762	Male	Has relevent experience	no_enrollment	Graduate	STEM	13	<10	Pvt Ltd	>4	18	1.0
8	27107	city_103	0.920	Male	Has relevent experience	no_enrollment	Graduate	STEM	7	50-99	Pvt Ltd	1	46	1.0
9	699	city_103	0.920	NaN	Has relevent experience	no_enrollment	Graduate	STEM	17	10000+	Pvt Ltd	>4	123	0.0

 

In [127]:

hr_df.isnull().sum()

 

Out[127]:

enrollee_id                  0
city                         0
city_development_index       0
gender                    4508
relevent_experience          0
enrolled_university        386
education_level            460
major_discipline          2813
experience                  65
company_size              5938
company_type              6140
last_new_job               423
training_hours               0
target                       0
dtype: int64

 

In [128]:

df_clean = hr_df.dropna()

 

In [129]:

df_clean.head(10)

 

Out[129]:

	enrollee_id	city	city_development_index	gender	relevent_experience	enrolled_university	education_level	major_discipline	experience	company_size	company_type	last_new_job	training_hours	target
1	29725	city_40	0.776	Male	No relevent experience	no_enrollment	Graduate	STEM	15	50-99	Pvt Ltd	>4	47	0.0
4	666	city_162	0.767	Male	Has relevent experience	no_enrollment	Masters	STEM	>20	50-99	Funded Startup	4	8	0.0
7	402	city_46	0.762	Male	Has relevent experience	no_enrollment	Graduate	STEM	13	<10	Pvt Ltd	>4	18	1.0
8	27107	city_103	0.920	Male	Has relevent experience	no_enrollment	Graduate	STEM	7	50-99	Pvt Ltd	1	46	1.0
11	23853	city_103	0.920	Male	Has relevent experience	no_enrollment	Graduate	STEM	5	5000-9999	Pvt Ltd	1	108	0.0
12	25619	city_61	0.913	Male	Has relevent experience	no_enrollment	Graduate	STEM	>20	1000-4999	Pvt Ltd	3	23	0.0
15	6588	city_114	0.926	Male	Has relevent experience	no_enrollment	Graduate	STEM	16	10/49	Pvt Ltd	>4	18	0.0
20	31972	city_159	0.843	Male	Has relevent experience	no_enrollment	Masters	STEM	11	100-500	Pvt Ltd	1	68	0.0
21	19061	city_114	0.926	Male	Has relevent experience	no_enrollment	Masters	STEM	11	100-500	Pvt Ltd	2	50	0.0
23	7041	city_40	0.776	Male	Has relevent experience	no_enrollment	Graduate	Humanities	<1	1000-4999	Pvt Ltd	1	65	0.0

 

In [131]:

sns.set_theme()
sns.set_context('paper')

g = sns.jointplot(data = df_clean, x = 'city_development_index', y = 'training_hours', hue = 'company_type', kind = 'kde')
g.fig.set_size_inches(15,15)

 

 

 

In [134]:

from google.colab import files
g.figure.savefig('seaborn_test.png')
files.download('seaborn_test.png')