Python में EDA — शुरुआती checklist
Read this guide in: English · मराठी · हिंदी
EDA मतलब सवालों के साथ table explore करना — बेतरतीब घूरना नहीं। pandas checklist: shape, head, info, missing, सादा groupby, फिर findings लिखो। Infosys-style hours: client A = 70.
दोस्तों! CSV खोलना EDA नहीं है। EDA = सवाल → inspect → note। आज shop, Instagram-like, Netflix-style या client sheets पर reuse होने वाली first-pass checklist। समझा?
संक्षिप्त उत्तर
खड़ा first-pass EDA:
- एक business सवाल लिखो।
read_csv→shape→head→columns.info()+isna().sum().- Numbers:
describe(); categories:value_counts. - सवाल का जवाब देने वाला एक groupby.
- Findings bullets + numbers (chart अगली guide में)।
question → inspect → clean notes → groupby → findings
Language: English · मराठी · हिंदी
इस guide से पहले क्या चाहिए?
- pandas load / filter / groupby / missing basics।
- Course: First-pass checklist.
Before और after
Before — Infosys-style client hours dump; no questions.
After — EDA checklist; top client A = 70 hours.
Before: raw hours dump। After: checklist भरी; top client A = 70.
EDA क्या है / नहीं है
है: सवाल‑guided looking, early quality issues, number-backed findings।
इस course पर नहीं: sklearn train, बिना वाक्य के ४० charts, cleaning हमेशा skip।
Infosys-style sample
| client | hours |
|---|---|
| A | 40 |
| B | NaN |
| A | 30 |
| C | 20 |
सवाल: किस client के सबसे ज़्यादा recorded hours?
Pass: shape (4,2) → hours में 1 NaN → decide fill/drop → A 70, C 20, B unknown → finding लिखो।
Checklist
Question → Load → Head/columns → Grain → Types → Missing → Duplicates → Univariate → groupby/crosstab → Findings → Next chart।
Correlation ≠ causation।
Good findings
- “Online shop sales total 190 (Mumbai 120, Pune 70).”
- “एक Netflix-style title पर missing watches; count KPI के लिए 0 fill।”
- “Client A = 70 hours; B incomplete।”
Drills
Shop channel → Online 190. Engagement city → Mumbai 120 vs Pune 70. Checklist वही — सवाल बदलता है।
Ravindra Bagale's Tip
Notebook के top Markdown में सवाल रखो। Cell उस सवाल की सेवा न करे तो हटाओ या नीचे ले जाओ। यही professional EDA। समझा?
Practice task
- सवाल एक वाक्य।
- shape, head, isna()।
- groupby client sum (non-missing)।
- तीन bullets — A = 70.
- अगला chart नाम (bar by client)।
Learn it properly
Related: Handle missing values · First charts
Samajla ka? EDA = question-driven pandas + लिखी findings — ML नहीं। Client A 70 finding है, vibe नहीं। अब आगे चलें।
Frequently asked questions
What is EDA?
Exploratory data analysis: structured looking before you claim insights or build dashboards.
Is EDA the same as cleaning?
Cleaning fixes types/missing/dupes. EDA uses clean-enough data to answer questions. They overlap in real notebooks.
Do I need seaborn for EDA?
Tables first. Charts help storytelling — covered in the visualisation guides/course chapters.
How long should first-pass EDA take?
For a small CSV, under an hour for a clear checklist. Depth comes from better questions.
Correlation caution?
High correlation is not proof of cause. The course has a caution lesson — read it.
Course links?
EDA chapter: question-driven EDA and first-pass checklist.