tidely 1.3.0b1

The Ultimate Data Cleaning Engine for Python

Tidely

The Operating System for Data Quality

Zero-Configuration • Explainable • Deterministic • Fast

---

Install

pip install tidely

The Magic

import tidely as td

result = td.clean("sales.csv")

clean_df = result.df

print(result.summary())

---

Why Tidely?

Real-world datasets are messy.

Missing values.

Broken dates.

Mixed datatypes.

Duplicate records.

Memory waste.

Encoding issues.

Schema drift.

Normally you spend hours writing cleaning scripts.

Tidely turns all of that into a single function call.

---

Dataset Intelligence

profile = td.inspect("sales.csv")

profile.show()

Output

✔ Trust Score

✔ Dataset DNA

✔ Semantic Detection

✔ Missing Values

✔ Duplicate Analysis

✔ Memory Analysis

✔ ML Readiness

✔ Data Quality Score

---

Why Use Tidely?

Feature	Pandas	Tidely
Read CSV	✅	✅
Auto Detect Dates	❌	✅
Auto Clean Dataset	❌	✅
Memory Optimization	Manual	Automatic
Duplicate Detection	Manual	Automatic
Missing Value Strategy	Manual	Automatic
Semantic Column Detection	❌	✅
Explain Every Change	❌	✅
Health Score	❌	✅
Trust Score	❌	✅
Production Summary	❌	✅

---

Production Validation

Tidely has been validated on

Dataset Type	Status
CSV	✅
Excel (.xlsx)	✅
ARFF	✅
Government Open Data	✅
Educational Data	✅
ML Benchmark Datasets	✅
Large CSV (>3 Million Rows)	✅
Time Series	✅
Mixed Datatypes	✅
Corrupted Data	✅

---

Validation Results

Version

v1.3.0-beta

Dataset	Rows	Health Before	Health After
Parking Meters	52	94	96
Credit-G	1000	86	90
Diabetes	768	86	92
Iris	150	92	92
Allegations	57	95	92
Mathematics	59	97	94

---

Benchmarks

3,055,000 Row Dataset

Metric	Result
Runtime	2.37 sec
Original Memory	148 MB
Final Memory	58 MB
Memory Saved	61%

---

Supported Formats

• CSV

• Excel

• Parquet

• JSON

• TSV

• Feather

• ARFF

More coming soon.

---

Explainable Cleaning

Tidely never silently changes your data.

Every transformation is documented.

Example

✓ Converted "Order Date" to datetime

Reason

Detected temporal values.

Impact

Allows time-series operations.

---

✓ Downcasted int64 → int16

Reason

Values fit inside Int16.

Impact

61% lower memory.

---

Philosophy

Tidely follows three principles.

Never silently modify data.

Every transformation is visible.

Deterministic.

Same input.

Same output.

Every time.

Local First.

Runs entirely on your machine.

No cloud.

No API keys.

No LLMs.

---

Roadmap

• CSV Cleaning

• Explainable Reports

• Memory Optimization

• Semantic Detection

• ARFF Support

• Excel Support

• Intelligent Missing Value Imputation

• Fuzzy Duplicate Detection

• Streaming Engine

• DuckDB Integration

• Out-of-Core Cleaning

• Auto Feature Engineering

• SQL Dataset Support

• Distributed Processing

---

Contributing

PRs are welcome.

Bug reports are welcome.

Feature requests are welcome.

---

License

MIT