Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Sign up for free
Menu
Search
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Pricing
Search
Sign in
Sign up for free
SKL 2019 Intern Training Python Data Analysis
Search
Sponsored
·
SiteGround - Reliable hosting with speed, security, and support you can count on.
→
Adam
March 05, 2019
Programming
59
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
SKL 2019 Intern Training Python Data Analysis
Data analysis and visualization introduction for 2019 intern.
Adam
March 05, 2019
More Decks by Adam
See All by Adam
Working Backward Reading Group
adamchang
0
140
Python Data Visualization - PyData Taipei Meetup
adamchang
0
250
SKL data analysis internship lecture 1
adamchang
0
160
SKL 2019 Intern Training Data Cleaning and Feature Engineering
adamchang
0
38
Other Decks in Programming
See All in Programming
Web Framework Benchmark Battle 2026, Any%
a_matsuda
2
2.1k
AI時代のコードレビューは人に向けるな、仕組みに向けろ
texmeijin
5
3.7k
re:Inventに行く前に知っておきたい現地参加のノウハウ
nokomoro3
0
430
AmebaブログはKMPから卒業します
kmars0605
0
190
The Rails Doctrine Decade
koic
2
590
當我們開始不讀 Code - AI Agent 時代,Code Review 與 Clean Code 還剩下什麼?
jame2408
1
210
パスワード再設定メールのURLについて知っておこう
uyuki234
0
160
UPDATE をやめる — EF Core でマスタをバージョン管理する
panda728
PRO
0
1.2k
C#の現在地 進化の歴史と、AI時代の.NET Everywhere
neuecc
5
5.2k
Unity6 URPによる海底をテーマにしたVJシステムの解説 | Unity Shader 完全に理解した 勉強会
gam0022
0
370
ActionController meets Protocol Buffers
sorah
0
520
[Rails World 2026] Durable orchestration on Rails: from continuation to workflow
palkan
1
750
Featured
See All Featured
HDC tutorial
michielstock
2
960
Statistics for Hackers
jakevdp
799
230k
[Rails World 2026] Durable orchestration on Rails: from continuation to workflow
palkan
1
750
brightonSEO & MeasureFest 2025 - Winning Strategies for Black Friday CRO & PPC - Christian Goodrich
cargoodrich
3
870
StorybookのUI Testing Handbookを読んだ
zakiyama
31
7k
Jamie Indigo - Trashchat’s Guide to Black Boxes: Technical SEO Tactics for LLMs
techseoconnect
PRO
0
710
Art, The Web, and Tiny UX
lynnandtonic
304
22k
Producing Creativity
orderedlist
PRO
348
41k
Design of three-dimensional binary manipulators for pick-and-place task avoiding obstacles (IECON2024)
konakalab
0
620
Design in an AI World
tapps
1
350
Test your architecture with Archunit
thirion
2
2.4k
GitHub's CSS Performance
jonrohan
1033
470k
Transcript
Data Analysis In Python 資料分析介紹 & Python實作
Yesterday recap! ‣ IDE In Python Development ‣ Markdown ‣
Data type ‣ Control Flow ‣ Exception ‣ Function ‣ Class ‣ Module ‣ Python Debugger
先介紹今天推薦的教材
為什麼要做資料分析? 幫助我們做更好的決策 - 我說的
–列⼦‧說符 「是故聖⼈⾒出以知入,觀往⽽知來,此其所 以先知之理也。」
TODAY’S AGENDA 1. Exploratory Data Analysis 2. Data Visualization 3.
How To Read Package Documents? 4. Pandas In Data Analysis 5. Homework
你會如何評估⼀份數據資料? 10 筆資料 100 筆資料 1,000,000 筆資料 平均值 平均值 平均值?????
分佈、分佈、分佈!
描述性統計 Min 25% Mediam 75% Max Min 12.5% 25% Mediam
75% 87.5% Max Min 12.5% 25% Mediam 75% 87.5% Max Midhinge Mideight Min 12.5% 25% Mediam 75% 87.5% Max Midhinge Mideight Counts Mean STD
Null Data
Outlier
Correlation
Data Visualization
– 分佈、分佈、分佈!!! 為什麼我們要做資料視覺化?
國內各⼈壽公司近⼀年在Google的搜索熱度 搜索熱度(數量) 0 25 50 75 100 搜索時間(⽇期) 2018-03-04 2018-03-25
2018-04-15 2018-05-06 2018-05-27 2018-06-17 2018-07-08 2018-07-29 2018-08-19 2018-09-09 2018-09-30 2018-10-21 2018-11-11 2018-12-02 2018-12-23 2019-01-13 2019-02-03 新光⼈壽: (Taiwan) 國泰⼈壽: (Taiwan) 富邦⼈壽: (Taiwan) 台灣⼈壽: (Taiwan) 我們如何看圖表? 1. 先看標頭 2. 座標軸名稱 3. 座標軸單位 4. 座標軸範圍 5. 看圖表Pattern & 圖⽰
常⽤的圖表有哪些?
Line Chart 0 25 50 75 100 4 ⽉ 5
⽉ 6 ⽉ 7 ⽉ Pros: 通常⽤於評估因時間變化 ⽽改變的趨勢 Cons: 結構簡單 通常需結合其他資訊比對
Bar Chart 0 25 50 75 100 4 ⽉ 5
⽉ 6 ⽉ 7 ⽉ Pros: 通常⽤於評估事件的頻率 Cons: 結構簡單 通常需結合其他資訊比對
Pie Chart 7% 8% 10% 11% 29% 35% Pros: 通常⽤於評估事件的佔比
Cons: 結構簡單 通常需結合其他資訊比對
Histogram Pros: 通常⽤於評估事件的頻率 分佈,是統計上常⽤的圖 表之⼀ Cons: 需調整資料級距 無法看到關鍵指標的位置
Scatter Plot Pros: 通常⽤於評估兩個因⼦之 間的關係 Cons: 常輔以回歸線看彼此的關 係
Boxplot Pros: 也是看分佈常⽤的圖標之 ⼀,可以清楚地看到描述 性統計 Cons: 評估整體性稍為直⽅圖弱 ⼀點,可以交叉比對看
Heatmap Pros: 通常⽤以看因⼦之間的相 關強度 Cons: ⽤於可明顯評估強度的資 料
How to read package document?
Python in jupyter
Recap • How to analyze data • How to visualize
data and get insight • How to use documentation • How to explore data using python
Introduce Kaggle https://www.kaggle.com
Homework - 資料探索松! Objective: 練習如何使⽤Python進⾏資料探勘,及對數據建立批判性思維 Dataset: Boston House Dataset From
Kaggle 1. 請針對各(78)個特徵做EDA,看看他們的描述性統計(如下表!) 2. 去看看每⼀個特徵跟房價是否有關係?請從中挑出15個你覺得有關係的特徵,並 挑出10個你覺得沒有關係的特徵,寫下你這樣判斷的理由 3. 將上述的資訊建成⼀張Excel的表,星期四驗收 Feature Chinese Data Descripti on Null Qty DataType Unit Mode Frequenc y(%) Counts Mean STD Min 25% Median 75% Max Relations hip? Reason GrLivAre a ⽣活區域 ⾯積 ⾯積⼤⼩ 100 Numerical m^2 35872 35 1.36 15 25 35 45 50 Yes 買房⼀定 看室內坪 數啊 Neighbor hood 社區位置 描述住宅 的區域 3000 Categoric al qty A 75 32872 Yes 房⼦的位 置Hen重 要