analysis3054 0.2.1

Advanced time-series analytics and forecasting toolkit for commodity and power trading

EIA Band Plot & Time Series Forecasting

This package provides two primary utilities:

• five_year_plot – Generate interactive 5‑year band plots using Plotly. These plots mirror the charts used by the U.S. Energy Information Administration (EIA) to contextualize recent values against the range, minimum, maximum and average of the last five years. Multiple numeric columns within a DataFrame can be plotted simultaneously as separate subplots.

• ml_forecast – Train individual AutoGluon time series models for each numeric column in a DataFrame and forecast future values. The function returns a DataFrame with point forecasts and, if requested, prediction intervals. Each series is trained independently using the specified presets (default: best_quality).

Installation

Install the package with:

pip install analysis3054

To enable the optional machine‑learning forecasting features, also install the AutoGluon time series dependency:

pip install analysis3054[ml]

Usage

Five‑Year Band Plot

import pandas as pd
from analysis3054 import five_year_plot

# Example DataFrame with a 'date' column and one or more numeric columns
df = pd.read_csv("my_timeseries_data.csv")

# Create the plot
fig = five_year_plot(date='date', df=df, prior_year_lines=1)
fig.show()

Machine Learning Forecasting

import pandas as pd
from analysis3054 import ml_forecast

df = pd.read_csv("my_timeseries_data.csv")

# Forecast the next 12 periods for each numeric column
result = ml_forecast(date='date', df=df, periods=12)

# Access point forecasts
forecasts = result.forecasts

# Access confidence intervals (if requested)
conf_ints = result.conf_intervals

See the docstrings of each function for detailed parameter descriptions.

User Guide

For a complete overview of all available functions, advanced forecasting methods, statistical analyses and plotting utilities, consult the USER_GUIDE.md file included with the package. It provides step‑by‑step examples, explains optional parameters such as confidence interval computation and plotting, and offers best practices for combining models and interpreting results.