edazer 0.2.0

lightweight library that provides functionalities for common EDA tasks

🚀 edazer

edazer is a lightweight Python package designed to accelerate exploratory data analysis (EDA) workflows. It provides simple, intuitive, and consistent APIs to inspect, summarize, and understand datasets—supporting both pandas and polars backends.

Instead of rewriting repetitive EDA code for every project, edazer helps you get insights in just a few lines.

---

📓 Kaggle Tutorial

👉 Quick hands-on guide:
https://www.kaggle.com/code/adarsh79x/edazer-for-quick-eda-pandas-polars-profiling

---

✨ What’s New in v0.2.0

• Improved backend abstraction for pandas & polars
• Cleaner API for dtype-based column selection
• Enhanced unique value inspection
• Better handling of edge cases (non-hashable columns, dtype normalization)
• Internal performance and structure improvements

---

🎯 Use Cases

• ⚡ Quick dataset understanding
• 📊 Early-stage data exploration
• 📓 Jupyter notebook workflows
• 🔍 Identifying data quality issues
• 🧠 Feature understanding before modeling

---

🔧 Features

📌 DataFrame Summary

Get a complete overview in one call:

• Schema / info
• Descriptive statistics
• Null percentages
• Duplicate count
• Unique values
• Shape

dz.summarize_df()

---

🔍 Smart Data Inspection

dz.lookup("head")     # first rows
dz.lookup("tail")     # last rows
dz.lookup("sample")   # random sample

---

🧩 Unique Value Exploration

dz.show_unique_values(
    column_names=["col1", "col2"],
    max_unique=10
)

• Automatically skips noisy columns
• Suggests when to increase threshold

---

🧠 Dtype-Based Column Selection

dz.cols_with_dtype(["float", "int"])

Options:

• exact=True → strict dtype match (float64)
• return_dtype_map=True → returns {column: dtype}

---

🔑 Primary Key Detection

from edazer import get_primary_key

get_primary_key(df, threshold=0.9, n_combos=2)

Find:

• Single-column unique identifiers
• Multi-column composite keys

---

📊 Data Profiling (Optional)

from edazer.profiling import show_data_profile

show_data_profile(dz)

Powered by ydata-profiling.

---

🖱️ Interactive Tables

from edazer import interactive_df

interactive_df()

Enables rich DataFrame viewing using itables.

---

📦 Installation

pip install edazer==0.2.0

---

⚡ Quick Start

import seaborn as sns
from edazer import Edazer

# Load dataset
df = sns.load_dataset("titanic")

# Initialize
dz = Edazer(df, backend="pandas")

# Summary
dz.summarize_df()

# Unique values
dz.show_unique_values(column_names=["sex", "class"])

# Dtype filtering
print(dz.cols_with_dtype(["float"]))

# Inspect data
dz.lookup("head")

---

📘 API Reference

Edazer(df, backend="pandas")

Create an analyzer instance.

• df: pd.DataFrame or pl.DataFrame
• backend: "pandas" or "polars"

---

summarize_df()

Displays:

• Schema/info
• Descriptive stats
• Null/duplicate counts
• Unique values
• Shape

---

show_unique_values(column_names, max_unique=10)

• column_names: list of columns
• max_unique: max values to display

---

cols_with_dtype(dtypes=None, exact=False, return_dtype_map=False)

• dtypes: list of dtype strings
• exact: strict match
• return_dtype_map: return dict instead of list

---

lookup(option="head")

• "head" → first rows
• "tail" → last rows
• "sample" → random rows

---

get_primary_key(df, threshold=0.9, n_combos=1, valid_column_dtypes=None)

Detect candidate keys.

Returns:

• List[str] or List[List[str]]

---

📊 Example Output

dz.show_unique_values(
    column_names=dz.cols_with_dtype(["object"])
)

sex: ['male', 'female']
embarked: ['S', 'C', 'Q', nan]
class: ['Third', 'First', 'Second']

---

🤝 Contributing

Contributions are welcome!

GitHub: https://github.com/adarsh-79/edazer

---

📄 License

MIT License

---

👨‍💻 Author

adarsh3690704