We have developed smoothDE, a density estimator that uses smoothing constraint to fit probability distributions with better Kullback-Leibler divergences than traditional density estimators. This smooth density estimator can be used to visualize data, infer sparsely populated data subsets, or improve machine learning performance.