#data-science

Linear least-squares regression as an API, computed locally and deterministically. The linear endpoint fits the best straight line y = a + b·x through a set of x/y data points by ordinary least squares, returning the slope b = Σ((x−x̄)(y−ȳ))/Σ(x−x̄)², the intercept a = ȳ − b·x̄, the ready-to-use equation, the Pearson correlation r and the coefficient of determination R² (the fraction of variance the line explains), and the residual and slope standard errors — the points (1,2),(2,4),(3,5),(4,4),(5,5) fit to y = 2.2 + 0.6·x with R² = 0.6, and a perfectly linear set returns R² = 1. Pass a predict_x and it also extrapolates the fitted value at that point. The predict endpoint evaluates y = intercept + slope·x for a known line. The x and y lists may be given as comma-separated values (x=1,2,3&y=2,4,5) or as JSON arrays in a POST body and must be equal length. Everything is computed locally and deterministically, so it is instant and private. Ideal for data-science, analytics, BI, forecasting, machine-learning-preprocessing and statistics-education app developers, trend-line and best-fit tools, and dashboards. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 2 endpoints. This is the regression line; for the Pearson correlation alone or descriptive statistics use a statistics API and for probability distributions a probability API.

api.oanor.com/regression-api

Classifier-evaluation maths as an API, computed locally and deterministically. The confusion endpoint turns the four cells of a binary confusion matrix — true and false positives and negatives — into the full metric suite: accuracy, precision, recall (sensitivity), specificity, the F1 score, the Matthews correlation coefficient (robust to class imbalance), balanced accuracy, negative predictive value, the false-positive and false-negative rates and the prevalence. The diagnostic endpoint applies Bayes' theorem to a medical or screening test: from its sensitivity, specificity and the prevalence (pre-test probability) it gives the positive and negative predictive values, the positive and negative likelihood ratios and the diagnostic odds ratio. The fbeta endpoint computes the Fβ score from precision and recall (or from the raw counts) for any β — β = 1 is F1, larger β weights recall, smaller β weights precision. Metrics whose denominator is zero are returned as null rather than erroring. Everything is computed locally and deterministically, so it is instant and private. Ideal for machine-learning, data-science, medical-testing and analytics app developers, model-evaluation and screening tools, and statistics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is classifier evaluation; for descriptive statistics and regression use a statistics API and for hypothesis tests an inference API.

api.oanor.com/classifier-api

The R package ecosystem — CRAN, the Comprehensive R Archive Network — as an API. Look up any R package for its title, description, version, license, maintainer and author, homepage and bug-tracker links, and its full dependency tree (depends, imports, suggests, linkingTo); read a package's complete release history with publication dates; search the entire CRAN registry by keyword; and get download statistics (last day, week or month, with an optional daily series) straight from the official CRAN download logs. Covers the ~22,000 packages on CRAN, from ggplot2, dplyr and data.table to jsonlite, shiny and the wider tidyverse. Live from the official R-community services (crandb, search.r-pkg.org, cranlogs). Ideal for package dashboards, dependency and supply-chain tooling, data-science developer portals and R ecosystem analytics. Open data from CRAN.

api.oanor.com/cran-api