[{"data":1,"prerenderedAt":12},["ShallowReactive",2],{"stack-tr-en-prophet":3},{"tagline":4,"description":5,"strengths":6,"useCase":11},"Meta's forecasting library that decomposes trend, seasonality and calendar effects.","Prophet fits an additive model in which the series is the sum of a trend (linear or logistic, with automatically detected changepoints), yearly, weekly and daily seasonalities modeled as Fourier series, and holiday or event effects supplied by the user. Bayesian fitting through Stan yields uncertainty intervals, tolerates missing values and outliers, and is tuned in business terms (capacity, changepoints, holiday calendar). It's our choice for a robust, explainable forecast of daily business series that can be up and running within hours.",[7,8,9,10],"Additive decomposition: piecewise trend, Fourier seasonalities, holiday effects.","Robust to missing values and outliers, no regular resampling required.","Bayesian uncertainty intervals via Stan, external regressors and capacity-bounded growth.","Parameters business analysts can read, built-in cross-validation, Python and R.","Daily traffic and sales forecasting, capacity planning, budget projections.",1792675980090]