You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
"Implemented two high-performance versions of the linear-model marginal likelihood in C/C++: one using LAPACKE and one using GSL. I engineered GEMM, identity-add, solve, and log-det computations with careful memory ownership and row/column-major handling. The implementations matched an R baseline and passed the spec check for a known subset.",
"Simulated repeated random samples to study the sampling distribution of proportions using the infer and tidyverse packages. Estimated p-hat for beliefs about scientists’ work benefiting society and visualized 15,000 samples with histograms. Compared sampling distributions at n = 10, 50, and 100, showing that as sample size increases, the shape becomes more normal, the mean approaches the true population proportion (0.2), and the standard error decreases.",
"Analyzed Youth Risk Behavior Survey (YRBSS) data to estimate the proportion of high schoolers who text while driving. Computed 99% and 95% confidence intervals for the true proportion using the infer package and visualized the margin of error as a function of population proportion. Conducted hypothesis tests (H₀: p=0.05) and found a p-value ≈ 0.00014, providing strong evidence that more than 5% of high schoolers text while driving.",
"Used the Youth Risk Behavior Survey (YRBSS) to test whether physically active high schoolers (≥3 days/week) weigh more than inactive peers. Created boxplots comparing groups, verified inference conditions, and ran two-sample t-tests using infer. Obtained a p-value ≈ 0.0002 for a two-sided test and 0.0001 for a one-sided alternative, leading to rejection of the null hypothesis at the 5% significance level.",
0 commit comments