E-CodeTI-89 Titanium statistics programs
9 checked statistics programs for the TI-89 Titanium. Each one runs in E-Code's virtual calculator in your browser, and you can put it on your own calculator.
The programs
- One-variable statistics of a list: Type a list like {2,4,4,5}; shows n, mean, median, sample standard deviation, min and max.
- Linear regression with exact slope: Least-squares line through lists of X and Y: slope and intercept kept exact when the data are exact (7/5, not 1.4), correlation r and r², and a prediction at any X.
- One-sample z test and interval (dialog): Two dialog boxes take the hypothesized mean, sigma, sample mean, n and alpha, and a drop-down picks the alternative; shows z, the p-value, the decision, and the confidence interval — normal tails by your own Define'd function.
- t test and interval from data lists: One-sample or pooled two-sample t test straight from your data lists (drop-down): t, degrees of freedom, two-sided p-value, the decision at alpha, t* and the confidence interval — the t distribution is your own Define'd function (exact finite sums for whole-number df).
- Chi-square goodness of fit: Observed counts and expected counts (or proportions) as lists: chi-square, degrees of freedom, the p-value from your own Define'd function (exact series for whole-number df), and a warning when an expected count is under 5.
- Binomial (exact), Poisson and normal probabilities: Pick a distribution from a pop-up: binomial P(X=k) and P(X≤k) as exact fractions when p is a fraction (105/512), Poisson P(X=k) and P(X≤k), or the normal probability between two values by numeric integration, with mean and SD.
- Least-squares polynomial fit (exact): Fit a line, parabola or cubic to X and Y lists by least squares — the coefficients come out as exact fractions when the data are exact — with R² and a prediction.
- Least squares with matrices (Vandermonde): X and Y lists and a degree 1-3: builds the Vandermonde matrix V and solves (VᵀV)c = VᵀY — exact coefficients for exact data — with R² and a prediction.
- Markov chain: n steps and the steady state: Type a transition matrix (rows sum to 1 — type fractions like 1/3 to stay exact), a starting row vector and a number of steps: the probability of each state after n steps (S1, S2, …) and the steady state (SS1, SS2, …).