dive-for-data 0.1.1

DIVE — Data Insight and Visualization Engine

DIVE — Data Insight and Visualization Engine

dive-for-data provides Dive, a small pure-Python container for ordered numeric data with:

• Statistical summaries (mean, median, mode, stdev, variance, skewness, kurtosis, etc.)
• Quantiles and histograms (percentile, quartiles, iqr)
• Data transforms (z_scores, normalized, cumulative_sum, moving_average, diff, pct_change, sorted, clip, apply)
• Prediction engine (predict_next, predict_detail, linear, quadratic, holt, exponential, drift, newton, lagrange, seasonal, ensemble)
• Regression/Correlation analysis (correlation, covariance, regress_on)
• ASCII visualizations (histogram, sparkline, plot_ascii)
• Utility exports (to_list, to_dict)

---

Installation

pip install dive-for-data

Quickstart

from dive import Dive

# Create dataset
sales = Dive([100, 150, 120, 200, 180])
print(sales.mean())          # 150.0
print(sales.summary())

# Add new value
sales += 220
print(sales[-1])            # 220.0

# Predict next value (ensemble model)
print(sales.predict_next())

# Use reference series for regression mode (len(reference) == len(self) + steps)
# This learns the relationship F(temps) -> sales
temps = Dive([20, 25, 22, 30, 28, 35])
print(sales.predict_next(reference=temps, TA=1))

# Detailed prediction report
report = sales.predict_detail(steps=3, reference=temps, TA=1)
print(report["ensemble"])

API overview

Data management

• Dive(data=None)
• add, append, remove, pop, clear, copy, data property
• supports Python protocols: len, indexing, iteration, in, equality

Stats

• mean, median, mode, geo_mean, harmonic_mean
• stdev, variance, range, min, max, sum
• percentile, quartiles, iqr.

Prediction

• predict_next(steps=1, method='ensemble', reference=None, corr_threshold=0.1, TA=0)
• predict_detail(...)

Cross-dataset

• correlation(other)
• covariance(other)
• regress_on(other)

Visualization

• summary() / describe()
• histogram(bins=10, width=40)
• sparkline()
• plot_ascii(width=60, height=15)

License

This project is licensed under the GNU General Public License v3 (GPLv3).