$devvkit learn --librarie pandas-guide
Pandas Guide
[python][data][analysis]
Python
Install
pip install pandas
Core data structure: DataFrame — a labeled, two-dimensional table with typed columns.
Handles real-world data messiness: NaN, inconsistent formats, duplicates.
Never loop over rows — use vectorized operations with apply(), groupby(), boolean indexing.
Creating DataFrames
Create DF— From dictionary.
import pandas as pd
df = pd.DataFrame({'name': ['Alice', 'Bob'], 'age': [30, 25]})Reading / Writing
Read CSV— Load CSV file.
df = pd.read_csv('data.csv')
df = pd.read_csv('data.csv', nrows=100)Write CSV— Save to CSV.
df.to_csv('output.csv', index=False)Selecting & Filtering
Inspect— View data.
df.head(10) df.info() df.describe()
Select columns— Choose columns.
df['name'] df[['name', 'age']]
Filter rows— Condition.
df[df['age'] > 25] df[(df['age'] > 25) & (df['city'] == 'NYC')]
Query— SQL-like filter.
df.query('age > 25 and city == 'NYC'')Manipulation
Add column— New column.
df['age_group'] = df['age'].apply(lambda x: 'young' if x < 30 else 'adult')
Merge— Join DataFrames.
pd.merge(df1, df2, on='user_id', how='left')
Grouping & Aggregation
Group by— Aggregate.
df.groupby('city').agg({'age': ['mean', 'count', 'max']})