AaronNHorvitz

AdaptiveLASSO

This is a modification of Scikit-Learn's LASSO for Adaptive LASSO to use for feature selection.

AIA-s-Introduction-to-Python

This is the first Python course I wrote for AIA Global's members (Another updated course is available.)

Federal-Election-Commission-FEC-Data

This is my capstone project. The full presentation can be found in the uploaded PDF file. I created a model that calculates the ROI and predicts individual donations to election campaigns for the US House of Representatives using data reported to the Federal Election Commission (FEC).

Find-Stock-and-Commodity-Price-Trends

How to find stock and commodity price trends using LOWESS (Locally Weighted Scatterplot Smoothing)

US-Presidential-Election-538-Poll-Reader

Here is a Jupyter Notebook that reads the polling data for the 2020 Presidential election published by 538 on Github, sorts through it, and plots the polling results and trends at the national and state levels. Have fun!

Binary-Classifier-Model-Testor-Selector

This tests various classifier models on binary target data to see which one maximizes performance. It then returns all the models, model performance statistics as well the ROC curves and Precision Recall plots, as well performance plots against the number of estimators used.

Data-Science-Training

Introductory Course for Data Science Training

Time-Series-Diagnostics

This code provides a series of diagnostic plots for time series analysis. These include ACF and PACF plots, predicted vs actual, residual plots, differenced residual plots, and many more.

Time-Series-Stationarity-Test

This a test for stationarity. It takes the original data, differenced data, plots both and performs a Dickey-Fuller Test

time_series_lowess_smoother

This is just a quick function to produce a lowess smoother.

Automobile-Data-Modeling-Exercise

Here is a quick run-through on creating a linear regression model using two popular variable selection/shrinkage methods (LASSO and Elastic-Net), followed by backward stepwise selection to arrive at a parsimonious model.

Energy-Information-Administration-EIA-and-FRED-API-calls-using-Python

Read-COVID-19-Data-Into-a-Pandas-Dataframe-Inside-A-Jupyter-Notebook

This is for anyone who would like to quickly read the COVID-19 time series data into a Pandas dataframe inside a Jupyter Notebook. I posted the Python code on Github in a Jupyter Notebook. The raw data is updated daily by Johns Hopkins University. Hopefully, this helps.

Rwork

R repo

Stock-Price-Reader

This is a simple notebook that will download stock prices for you and display them on a chart.

Tests-For-Normality

COVID-19-Forecast-With-Exogenous

COVID-19 Forecast With Exogenous Variables

data

Data and code behind the articles and graphics at FiveThirtyEight

FECDataConnect

FECDataConnect is a project aimed at extracting data from the Federal Election Commission (FEC) and integrating it into a MariaDB database. This ETL pipeline ensures that the data remains fresh and accessible for further analysis.

gaminet

Generalized additive model with pairwise interactions

pyloess

A simple implementation of the LOESS algorithm using numpy

PythonDataAnalytics

Code for common data analytics projects

Raising-Money-for-Congressional-Elections-Capstone-Project

This is my capstone project. I created a model that calculates the ROI and predicts individual donations to election campaigns for the US House of Representatives using data reported to the Federal Election Commission (FEC).

Snippets

These are just small pieces of code that may help make your life easier.

Time-Series-Diagnostic-Plots

This code provides a series of diagnostic plots for time series analysis. These include ACF and PACF plots, predicted vs actual, residual plots, differenced residual plots, and many more.

Upsample-Binary-Classification-Data

Code to upsample binary classification data when you have a rare event problem.