<div id="readability-page-1" class="page"><div>

 “10 minutes to pandas” 문서를 따라해보며 작성했습니다.




<a href="https://pandas.pydata.org/pandas-docs/stable/user_guide/10min.html">https://pandas.pydata.org/pandas-docs/stable/user_guide/10min.html</a>








 2022. 1. 30 최초작성





 2022. 3. 19









 head에 지정한 개수만큼 DataFrame 상단의 행을 볼 수 있습니다. 디폴트 값은 5입니다. 여기에선 3으로 지정하고 있습니다. 




<pre><code>import numpy as np
import pandas as pd


df = pd.DataFrame(np.random.randn(20, 2), columns=list("AB"))

print(df.head(3))</code></pre>



<figure><img data-lazy-fallback="1" data-attachment-id="5640" data-permalink="https://webnautes.blog/5639/word-image-5639-1/" data-orig-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-1.png?fit=376%2C198&amp;ssl=1" data-orig-size="376,198" data-comments-opened="1" data-image-meta="{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}" data-image-title="word-image-5639-1" data-image-description="" data-image-caption="" data-medium-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-1.png?fit=300%2C158&amp;ssl=1" data-large-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-1.png?fit=376%2C198&amp;ssl=1" decoding="async" fetchpriority="high" width="376" height="198" src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-1.png?resize=376%2C198&amp;ssl=1" alt="" sizes="(max-width: 376px) 100vw, 376px" data-recalc-dims="1" data-old-src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-1.png?resize=376%2C198&amp;is-pending-load=1#038;ssl=1" data-lazy-srcset="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-1.png?w=376&amp;ssl=1 376w, https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-1.png?resize=300%2C158&amp;ssl=1 300w"></figure>








 tail에 지정한 개수만큼 DataFrame 하단의 행을 볼 수 있습니다. 디폴트 값은 5입니다. 여기에선 3으로 지정했습니다. 




<pre><code>import numpy as np
import pandas as pd


df = pd.DataFrame(np.random.randn(20, 2), columns=list("AB"))

print(df.tail(3))</code></pre>



<figure><img data-lazy-fallback="1" data-attachment-id="5641" data-permalink="https://webnautes.blog/5639/word-image-5639-2/" data-orig-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-2.png?fit=378%2C206&amp;ssl=1" data-orig-size="378,206" data-comments-opened="1" data-image-meta="{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}" data-image-title="word-image-5639-2" data-image-description="" data-image-caption="" data-medium-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-2.png?fit=300%2C163&amp;ssl=1" data-large-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-2.png?fit=378%2C206&amp;ssl=1" decoding="async" width="378" height="206" src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-2.png?resize=378%2C206&amp;ssl=1" alt="" sizes="(max-width: 378px) 100vw, 378px" data-recalc-dims="1" data-old-src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-2.png?resize=378%2C206&amp;is-pending-load=1#038;ssl=1" data-lazy-srcset="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-2.png?w=378&amp;ssl=1 378w, https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-2.png?resize=300%2C163&amp;ssl=1 300w"></figure>








 인덱스와 열 이름을 각각 출력합니다. 이때 데이터프레임의 values 속성을 사용하여 Numpy 배열로 변환했습니다.




<pre><code>import numpy as np
import pandas as pd


df = pd.DataFrame(np.random.randn(3, 3), columns=list("ABC"))

print(df)
print(df.index.values)
print(df.columns.values)</code></pre>



<figure><img data-lazy-fallback="1" data-attachment-id="5642" data-permalink="https://webnautes.blog/5639/word-image-5639-3/" data-orig-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-3.png?fit=488%2C302&amp;ssl=1" data-orig-size="488,302" data-comments-opened="1" data-image-meta="{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}" data-image-title="word-image-5639-3" data-image-description="" data-image-caption="" data-medium-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-3.png?fit=300%2C186&amp;ssl=1" data-large-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-3.png?fit=488%2C302&amp;ssl=1" decoding="async" width="488" height="302" src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-3.png?resize=488%2C302&amp;ssl=1" alt="" sizes="(max-width: 488px) 100vw, 488px" data-recalc-dims="1" data-old-src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-3.png?resize=488%2C302&amp;is-pending-load=1#038;ssl=1" data-lazy-srcset="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-3.png?w=488&amp;ssl=1 488w, https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-3.png?resize=300%2C186&amp;ssl=1 300w"></figure>








 to_numpy() 메소드 또는 values 속성을 사용하여 DataFrame을 넘파이 배열로 변환합니다. 





 데이터프레임에 속한 데이터의 타입이 모두 같지 않으면 넘파이 배열로 변환시 타입이 모두 같을 때에 비해 시간이 더 걸립니다.





 왜냐하면 넘파이 배열은 모두 같은 타입을 가지기 때문에 데이터 프레임의 데이터 타입이 컬럼별로 다른 경우 모든 타입을 하나의 데이터 타입으로 변환해야 하기 때문입니다.





 데이터프레임 df의 경우 모든 컬럼이 같은 타입이고 데이터프레임 df2는 컬럼별로 다른 타입으로 구성되어 있습니다. 





 df(0.0초)보다 df2(2.7초)가 넘파이 배열로 변환하는데 걸린 시간이 더 오래 걸린 것을 볼 수 있습니다.




<pre><code>import numpy as np
import pandas as pd
import time


df = pd.DataFrame(np.random.randn(1000000, 2), columns=list("AB"))

start = time.time() 

n1 = df.to_numpy()

print("time1 :", time.time() - start) 
# time1 : 0.0

print(n1.dtype)
# float64


df2 = pd.DataFrame(
 {
 "A": np.random.randn(1000000),
 "B": pd.Timestamp("20211211")
 }
)

start = time.time() 

n2 = df2.to_numpy()

print("time2 :", time.time() - start) 
# time2 : 2.714345932006836

print(n2.dtype)
# object</code></pre>








 to_numpy()를 사용하여 데이터프레임을 NumPy 배열로 변환한 경우 인덱스와 컬럼 라벨은 포함되지 않습니다. 




<pre><code>import numpy as np
import pandas as pd


df = pd.DataFrame(np.random.randn(4, 2), columns=list("AB"))
print(df)


print()


n1 = df.to_numpy()
print(n1)</code></pre>



<figure><img data-lazy-fallback="1" data-attachment-id="5643" data-permalink="https://webnautes.blog/5639/word-image-5639-4/" data-orig-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-4.png?fit=456%2C480&amp;ssl=1" data-orig-size="456,480" data-comments-opened="1" data-image-meta="{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}" data-image-title="word-image-5639-4" data-image-description="" data-image-caption="" data-medium-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-4.png?fit=285%2C300&amp;ssl=1" data-large-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-4.png?fit=456%2C480&amp;ssl=1" decoding="async" loading="lazy" width="456" height="480" src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-4.png?resize=456%2C480&amp;ssl=1" alt="" sizes="(max-width: 456px) 100vw, 456px" data-recalc-dims="1" data-old-src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-4.png?resize=456%2C480&amp;is-pending-load=1#038;ssl=1" data-lazy-srcset="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-4.png?w=456&amp;ssl=1 456w, https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-4.png?resize=285%2C300&amp;ssl=1 285w"></figure>








 describe()를 사용하면 데이터프레임에 저장된 데이터의 통계를 보여줍니다. 




<pre><code>import numpy as np
import pandas as pd


df = pd.DataFrame(np.arange(100))

print(df.describe())</code></pre>



<figure><img data-lazy-fallback="1" data-attachment-id="5644" data-permalink="https://webnautes.blog/5639/word-image-5639-5/" data-orig-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-5.png?fit=318%2C412&amp;ssl=1" data-orig-size="318,412" data-comments-opened="1" data-image-meta="{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}" data-image-title="word-image-5639-5" data-image-description="" data-image-caption="" data-medium-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-5.png?fit=232%2C300&amp;ssl=1" data-large-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-5.png?fit=318%2C412&amp;ssl=1" decoding="async" loading="lazy" width="318" height="412" src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-5.png?resize=318%2C412&amp;ssl=1" alt="" sizes="(max-width: 318px) 100vw, 318px" data-recalc-dims="1" data-old-src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-5.png?resize=318%2C412&amp;is-pending-load=1#038;ssl=1" data-lazy-srcset="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-5.png?w=318&amp;ssl=1 318w, https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-5.png?resize=232%2C300&amp;ssl=1 232w"></figure>








 데이터프레임을 전치(Transpose)합니다.




<pre><code>import numpy as np
import pandas as pd


df = pd.DataFrame([[1, 2], [3, 4]], columns=list("AB"))

print(df)
print()
print(df.T)</code></pre>



<figure><img data-lazy-fallback="1" data-attachment-id="5645" data-permalink="https://webnautes.blog/5639/word-image-5639-6/" data-orig-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-6.png?fit=192%2C330&amp;ssl=1" data-orig-size="192,330" data-comments-opened="1" data-image-meta="{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}" data-image-title="word-image-5639-6" data-image-description="" data-image-caption="" data-medium-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-6.png?fit=175%2C300&amp;ssl=1" data-large-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-6.png?fit=192%2C330&amp;ssl=1" decoding="async" loading="lazy" width="192" height="330" src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-6.png?resize=192%2C330&amp;ssl=1" alt="" sizes="(max-width: 192px) 100vw, 192px" data-recalc-dims="1" data-old-src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-6.png?resize=192%2C330&amp;is-pending-load=1#038;ssl=1" data-lazy-srcset="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-6.png?w=192&amp;ssl=1 192w, https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-6.png?resize=175%2C300&amp;ssl=1 175w"></figure>








 축(axis)을 기준으로 정렬합니다.





 첫번째 축을(axis=0) 따라 정렬하면 행을 기준으로 하며 ascending를 False로 했기 때문에 내림차순으로 정렬됩니다. 





 두번째 축을(axis=1) 따라 정렬하면 열을 기준으로 하며 ascending를 False로 했기 때문에 내림차순으로 정렬됩니다. 




<pre><code>import numpy as np
import pandas as pd


df = pd.DataFrame([[1, 2, 3, 4],[5, 6, 7, 8],[9, 10, 11, 12]], columns=list("ABCD"))


print(df)
print()

print(df.sort_index(axis=0, ascending=False))
print()

print(df.sort_index(axis=1, ascending=False))</code></pre>



<figure><img data-lazy-fallback="1" data-attachment-id="5646" data-permalink="https://webnautes.blog/5639/word-image-5639-7/" data-orig-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-7.png?fit=352%2C670&amp;ssl=1" data-orig-size="352,670" data-comments-opened="1" data-image-meta="{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}" data-image-title="word-image-5639-7" data-image-description="" data-image-caption="" data-medium-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-7.png?fit=158%2C300&amp;ssl=1" data-large-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-7.png?fit=352%2C670&amp;ssl=1" decoding="async" loading="lazy" width="352" height="670" src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-7.png?resize=352%2C670&amp;ssl=1" alt="" sizes="(max-width: 352px) 100vw, 352px" data-recalc-dims="1" data-old-src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-7.png?resize=352%2C670&amp;is-pending-load=1#038;ssl=1" data-lazy-srcset="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-7.png?w=352&amp;ssl=1 352w, https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-7.png?resize=158%2C300&amp;ssl=1 158w"></figure>








 특정 컬럼의 값을 기준으로 정렬합니다.





 아래 예제에서는 B 컬럼의 값이 뒤섞여 있는 상태인데 B 컬럼의 값이 내림차순으로 정렬되도록 하여 나머지 컬럼도 정렬되었습니다.




<pre><code>import numpy as np
import pandas as pd


df = pd.DataFrame([[1, 3, 3, 4],[5, 0, 7, 8],[9, 2, 11, 12]], columns=list("ABCD"))


print(df)
print()

print(df.sort_values(by="B"))</code></pre>



<figure><img data-lazy-fallback="1" data-attachment-id="5647" data-permalink="https://webnautes.blog/5639/word-image-5639-8/" data-orig-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-8.png?fit=302%2C410&amp;ssl=1" data-orig-size="302,410" data-comments-opened="1" data-image-meta="{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}" data-image-title="word-image-5639-8" data-image-description="" data-image-caption="" data-medium-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-8.png?fit=221%2C300&amp;ssl=1" data-large-file="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-8.png?fit=302%2C410&amp;ssl=1" decoding="async" loading="lazy" width="302" height="410" src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-8.png?resize=302%2C410&amp;ssl=1" alt="" sizes="(max-width: 302px) 100vw, 302px" data-recalc-dims="1" data-old-src="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-8.png?resize=302%2C410&amp;is-pending-load=1#038;ssl=1" data-lazy-srcset="https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-8.png?w=302&amp;ssl=1 302w, https://i0.wp.com/webnautes.blog/wp-content/uploads/2023/08/word-image-5639-8.png?resize=221%2C300&amp;ssl=1 221w"></figure>





</div></div>

멈춤보단 천천히라도

webnautes

“10 minutes to pandas” 문서를 따라해보며 작성했습니다. https://pandas.pydata.org/pandas-docs/stable/user_guide/10min.html 2022. 1. 30 최초작성 2022. 3. 19 head에 지정한 개수만큼 DataFrame 상단의 행을 볼 수 있습니다. 디폴트 값은 5입니다. 여기에선 3으로 지정하고 있습니다. tail에 지정한 개수만큼 DataFrame 하단의 행을 볼 수 있습니다. 디폴트 값은 5입니다. 여기에선 3으로 지정했습니다. 인덱스와 열 이름을 각각 출력합니다. 이때 데이터프레임의 values 속성을… 더 보기 »Pandas 강좌 2 – 데이터 보는 방법
The post Pandas 강좌 2 – 데이터 보는 방법 first appeared on 멈춤보단 천천히라도.

Pandas 강좌 2 – 데이터 보는 방법

댓글