Pandas¶

Pandas 는 엑셀과 비슷한 역할을 하는 Python의 주요한 라이브러리 입니다. 데이터를 읽고, 처리하고, 분석할 수 있습니다.

• Pandas 라이브러리를 이용해서 엑셀 형식의 파일 (Spreadsheet, ex. csv, xlsx, ...) 을 열어봅니다.
• DataFrame를 사용하는 법을 배워봅니다.

1. 파일 로드하고 살펴보기¶

• 경로 (Path) = 파일 위치 & 파일 이름

• / : root. Windows에서는 기본옵션에서 C:\
• ~/ : 사용자 폴더. Windows에서는 C:\Users\사용자계정이름
• ./ : 현재 작업 폴더(working directory), 별도로 작업하지 않은 경우 생략 가능
• ../ : 현재 폴더의 상위 폴더

• Windows 에는 "한글" 을 조심할 것

In [ ]:

# print working directory

%pwd

# print working directory %pwd

In [ ]:

# pandas 설치

%pip install -U pandas

# pandas 설치 %pip install -U pandas

In [12]:

import pandas as pd

df = pd.read_csv("./AAPL_data.csv")
df

import pandas as pd df = pd.read_csv("./AAPL_data.csv") df

Out[12]:

	Price	Close	High	Low	Open	Volume
0	Ticker	AAPL	AAPL	AAPL	AAPL	AAPL
1	Date	NaN	NaN	NaN	NaN	NaN
2	2023-01-03	123.7684555053711	129.53777972145332	122.87781986866916	128.92423659950236	112117500
3	2023-01-04	125.04503631591797	127.32110440506466	123.77835783326215	125.56951966733021	89113600
4	2023-01-05	123.71897888183594	126.44036106917035	123.46169000194207	125.80702181866339	80962700
...	...	...	...	...	...	...
247	2023-12-22	192.6561737060547	194.45734722393942	192.02924020237552	194.22845757962483	37122800
248	2023-12-26	192.10885620117188	192.94475743470915	191.88992751842636	192.66612369019674	28919300
249	2023-12-27	192.20835876464844	192.5566585360189	190.15840400250286	191.55158790358445	48087700
250	2023-12-28	192.6362762451172	193.71101293849657	192.22827139973785	193.1935437488545	34049900
251	2023-12-29	191.5913848876953	193.452263485871	190.7952819760843	192.9547010641641	42628800

252 rows × 6 columns

In [13]:

df = df[2:]
df

df = df[2:] df

Out[13]:

	Price	Close	High	Low	Open	Volume
2	2023-01-03	123.7684555053711	129.53777972145332	122.87781986866916	128.92423659950236	112117500
3	2023-01-04	125.04503631591797	127.32110440506466	123.77835783326215	125.56951966733021	89113600
4	2023-01-05	123.71897888183594	126.44036106917035	123.46169000194207	125.80702181866339	80962700
5	2023-01-06	128.27110290527344	128.93412872931725	123.59032994150238	124.6986773645302	87754700
6	2023-01-09	128.7955780029297	132.02166222689684	128.53828914886986	129.11225514638548	70790800
...	...	...	...	...	...	...
247	2023-12-22	192.6561737060547	194.45734722393942	192.02924020237552	194.22845757962483	37122800
248	2023-12-26	192.10885620117188	192.94475743470915	191.88992751842636	192.66612369019674	28919300
249	2023-12-27	192.20835876464844	192.5566585360189	190.15840400250286	191.55158790358445	48087700
250	2023-12-28	192.6362762451172	193.71101293849657	192.22827139973785	193.1935437488545	34049900
251	2023-12-29	191.5913848876953	193.452263485871	190.7952819760843	192.9547010641641	42628800

250 rows × 6 columns

In [ ]:

df.head()

df.head()

In [ ]:

df.tail(n=10)

df.tail(n=10)

In [ ]:

df.shape

df.shape

In [ ]:

df.index

df.index

In [ ]:

df.columns

df.columns

In [14]:

df.dtypes

df.dtypes

Out[14]:

Price     object
Close     object
High      object
Low       object
Open      object
Volume    object
dtype: object

2. 데이터 다루기¶

• 데이터 타입 설정하기
• 데이터 선택하기, 정렬하기
• 데이터 선택을 위한 조건 생성하고 그대로 사용하기

애플은 1년동안 며칠이나 상승을 했고, 며칠이나 하락을 했을까요?

In [ ]:

df.sample(n=10)
df.sample(frac=0.1)

df.sample(n=10) df.sample(frac=0.1)

In [24]:

df.loc[:, "Volume"] = df["Volume"].astype('float')

df.loc[:, "Volume"] = df["Volume"].astype('float')

In [19]:

df.nlargest(10,"Volume")

df.nlargest(10,"Volume")

Out[19]:

	Price	Close	High	Low	Open	Volume
24	2023-02-03	152.89219665527344	155.74223078414798	146.29160978319123	146.4895254643634	154357300.0
242	2023-12-15	196.6068115234375	197.43275173465906	196.0395831061416	196.5669980286803	128256700.0
107	2023-06-05	178.2287139892578	183.55830143833987	176.70029356551757	181.25576664565278	121946500.0
23	2023-02-02	149.25050354003906	149.60674271507267	146.62807162382762	147.35047067319957	118339000.0
149	2023-08-04	180.62060546875	185.97004732736383	180.5511249208849	184.12404245746146	115799700.0
87	2023-05-05	172.0260009765625	172.74950296676494	169.24098486047572	169.45902904213452	113316400.0
172	2023-09-07	176.46188354492188	177.10787273977496	172.46674085776903	174.0965977246818	112488800.0
2	2023-01-03	123.7684555053711	129.53777972145332	122.87781986866916	128.92423659950236	112117500.0
178	2023-09-15	173.92764282226562	175.40843335624928	172.74501513151833	175.38855280048267	109205100.0
116	2023-06-16	183.52854919433594	185.58298054193108	182.88344624177174	185.32492724572717	101235600.0

In [20]:

df.sort_values('Volume')

df.sort_values('Volume')

Out[20]:

	Price	Close	High	Low	Open	Volume
227	2023-11-24	189.0438690185547	189.96932784078396	188.3273779116158	189.93947531002817	24048300.0
248	2023-12-26	192.10885620117188	192.94475743470915	191.88992751842636	192.66612369019674	28919300.0
126	2023-07-03	191.0117950439453	192.4211080946957	190.3170502473471	192.32185451119005	31458200.0
250	2023-12-28	192.6362762451172	193.71101293849657	192.22827139973785	193.1935437488545	34049900.0
146	2023-08-01	194.1381072998047	195.24967486566203	193.81058861115622	194.7633716275856	35175100.0
...	...	...	...	...	...	...
149	2023-08-04	180.62060546875	185.97004732736383	180.5511249208849	184.12404245746146	115799700.0
23	2023-02-02	149.25050354003906	149.60674271507267	146.62807162382762	147.35047067319957	118339000.0
107	2023-06-05	178.2287139892578	183.55830143833987	176.70029356551757	181.25576664565278	121946500.0
242	2023-12-15	196.6068115234375	197.43275173465906	196.0395831061416	196.5669980286803	128256700.0
24	2023-02-03	152.89219665527344	155.74223078414798	146.29160978319123	146.4895254643634	154357300.0

250 rows × 6 columns

In [7]:

df["Close"]

df["Close"]

Out[7]:

0                    AAPL
1                     NaN
2       123.7684555053711
3      125.04503631591797
4      123.71897888183594
              ...        
247     192.6561737060547
248    192.10885620117188
249    192.20835876464844
250     192.6362762451172
251     191.5913848876953
Name: Close, Length: 252, dtype: object

In [ ]:

In [25]:

df.loc[:,"Close"] = df["Close"].astype('float')
df.loc[:,"Open"] = df['Open'].astype('float')
df['Close'] - df['Open']

df.loc[:,"Close"] = df["Close"].astype('float') df.loc[:,"Open"] = df['Open'].astype('float') df['Close'] - df['Open']

Out[25]:

2     -5.155781
3     -0.524483
4     -2.088043
5      3.572426
6     -0.316677
         ...   
247   -1.572284
248   -0.557267
249    0.656771
250   -0.557268
251   -1.363316
Length: 250, dtype: float64

In [26]:

df["Close"] > df['Open']

df["Close"] > df['Open']

Out[26]:

2      False
3      False
4      False
5       True
6      False
       ...  
247    False
248    False
249     True
250    False
251    False
Length: 250, dtype: bool

In [27]:

df[df["Close"] > df['Open']]

df[df["Close"] > df['Open']]

Out[27]:

	Price	Close	High	Low	Open	Volume
5	2023-01-06	128.271103	128.93412872931725	123.59032994150238	124.698677	87754700.0
7	2023-01-10	129.369583	129.89406659487034	126.78674290984911	128.904473	63896200.0
8	2023-01-11	132.100845	132.12062633545213	129.1023781585197	129.884150	69458900.0
10	2023-01-13	133.357620	133.51595882992456	130.28988932008468	130.656034	57809700.0
11	2023-01-17	134.525345	135.86128703391867	132.73418300243557	133.426895	63646600.0
...	...	...	...	...	...	...
240	2023-12-13	196.994919	197.03471713546693	193.90007998207173	194.138900	70404200.0
241	2023-12-14	197.144165	198.6467979467734	195.20367481138322	197.054607	66831600.0
242	2023-12-15	196.606812	197.43275173465906	196.0395831061416	196.566998	128256700.0
244	2023-12-19	195.979889	195.98983469805682	194.9350047944863	195.203693	40714100.0
249	2023-12-27	192.208359	192.5566585360189	190.15840400250286	191.551588	48087700.0

151 rows × 6 columns

In [ ]:

3. 데이터 집계하기¶

• 평균내고, 편차 구하고, 다양한 계산 행하기

In [28]:

df.describe()

df.describe()

Out[28]:

	Close	Open	Volume
count	250.000000	250.000000	2.500000e+02
mean	171.281995	170.992146	5.921703e+07
std	17.418788	17.615621	1.777392e+07
min	123.718979	124.698677	2.404830e+07
25%	160.670422	160.117881	4.781208e+07
50%	174.389793	174.161200	5.507750e+07
75%	186.265335	185.399351	6.574292e+07
max	197.144165	197.054607	1.543573e+08

In [31]:

df[["Close", "Open", "Volume"]].corr()

df[["Close", "Open", "Volume"]].corr()

Out[31]:

	Close	Open	Volume
Close	1.000000	0.994800	-0.321075
Open	0.994800	1.000000	-0.324514
Volume	-0.321075	-0.324514	1.000000

In [34]:

df["date"] = pd.to_datetime(df['Price'])
df['weekday'] = df['date'].dt.weekday
df

df["date"] = pd.to_datetime(df['Price']) df['weekday'] = df['date'].dt.weekday df

/var/folders/kx/c6xk17ln6blbs7f3nf_p1str0000gn/T/ipykernel_8397/3916596524.py:1: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df["date"] = pd.to_datetime(df['Price'])
/var/folders/kx/c6xk17ln6blbs7f3nf_p1str0000gn/T/ipykernel_8397/3916596524.py:2: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df['weekday'] = df['date'].dt.weekday

Out[34]:

	Price	Close	High	Low	Open	Volume	date	weekday
2	2023-01-03	123.768456	129.53777972145332	122.87781986866916	128.924237	112117500.0	2023-01-03	1
3	2023-01-04	125.045036	127.32110440506466	123.77835783326215	125.569520	89113600.0	2023-01-04	2
4	2023-01-05	123.718979	126.44036106917035	123.46169000194207	125.807022	80962700.0	2023-01-05	3
5	2023-01-06	128.271103	128.93412872931725	123.59032994150238	124.698677	87754700.0	2023-01-06	4
6	2023-01-09	128.795578	132.02166222689684	128.53828914886986	129.112255	70790800.0	2023-01-09	0
...	...	...	...	...	...	...	...	...
247	2023-12-22	192.656174	194.45734722393942	192.02924020237552	194.228458	37122800.0	2023-12-22	4
248	2023-12-26	192.108856	192.94475743470915	191.88992751842636	192.666124	28919300.0	2023-12-26	1
249	2023-12-27	192.208359	192.5566585360189	190.15840400250286	191.551588	48087700.0	2023-12-27	2
250	2023-12-28	192.636276	193.71101293849657	192.22827139973785	193.193544	34049900.0	2023-12-28	3
251	2023-12-29	191.591385	193.452263485871	190.7952819760843	192.954701	42628800.0	2023-12-29	4

250 rows × 8 columns

In [38]:

df.groupby("weekday")[['Open', 'Close', 'Volume']].mean()

df.groupby("weekday")[['Open', 'Close', 'Volume']].mean()

Out[38]:

	Open	Close	Volume
weekday
0	171.471788	171.955209	5.626837e+07
1	170.407503	170.581198	5.477188e+07
2	170.997955	170.949820	5.886871e+07
3	170.648236	170.965807	6.017473e+07
4	171.491563	172.043658	6.566139e+07