- Create templates for writing future programs related to specific statistic methods and models
- Templates written in R coding language
- Created in jupyter notebook with markdown
- Content:
- Check ANOVA model assumptions
- Create ANOVA equations and tables
- One Way ANOVA/CRD Model, Two Way ANOVA/RCBD Model, Two Way Full Model with Interaction, Latin Square Incomplete Block Design, 2^k Factorial Design, (2^N, 2^k) Design
- Least squares estimates of parameters
- Hypothesis tests and confidence intervals for contrasts and multiple comparisons
- Content:
- Check linear regression assumptions
- Transform variables with Invariance Transformation, Power Transformation, and Box-Cox Transformation
- Identify outliers, leverage points, and influential points
- Create multiple linear regression and polynomial regression models for numeric and categorical data
- Hypothesis tests for significant predictors, significant categorical variables, significant interaction effects, and significant model degrees
- Predictor selection with Stepwise Regression and Regression Subsets
- Least squares estimates and weighted least squares estimates of coefficients
- Confidence intervals of new responses
- Content:
- Simulating ARMA models
- Stabilizing variance with Box-Cox
- Differencing to remove seasonality and trend
- Calculate Autocovariance Functions, Autocorrelation Functions, and Partial Autocorrelation Functions for time series data
- Model identification of SARIMA(P,D,Q)x(p,d,q)s using Sample Autocorrelation and Sample Partial Autocorrelation Functions
- Comparing different models with AICC
- Preliminary coefficient estimates with Yule-Walker Equations, Durbin-Levinson Algorithm, and Innovation Algorithm
- Maximum likelihood estimated coefficients
- Diagnostic checking
- Forecasting
Thank you to my professors and TAs at UCSB in PSTAT 122, PSTAT 126, and PSTAT 174 for teaching me these coding concepts.