A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Python artificial-intelligencedagdata-sciencedata-visualizationdataflowdeveloper-toolsmachine-learningnotebookspipelinepythonreactivesqlweb-app
/* Copyright 2026 Marimo. All rights reserved. */
import { Logger } from "./Logger";
export const Maps = {
/**
* keyBy for Map, with duplicate key detection.
*/
keyBy<V, K = string>(
items: Iterable<V>,
key: (item: NoInfer<V>) => K,
): Map<K, V> {
const map = new Map<K, V>();
const duplicateIds = new Set<K>();
for (const item of items) {
const k = key(item);
if (map.has(k)) {
duplicateIds.add(k);
}
map.set(k, item);
}
if (duplicateIds.size > 0) {
Logger.trace(`Duplicate keys: ${[...duplicateIds].join(", ")}`);
}
return map;
},
collect<T, V, K = string>(
items: Iterable<T>,
key: (item: NoInfer<T>) => K,
mapper: (item: NoInfer<T>) => V,
): Map<K, V> {
return Maps.mapValues(Maps.keyBy(items, key), mapper);
},
/**
* Filter a map by a predicate.
*/
filterMap<K, V>(
map: Map<K, V>,
predicate: (value: V, key: K) => boolean,
): Map<K, V> {
const result = new Map<K, V>();
for (const [key, value] of map) {
if (predicate(value, key)) {
result.set(key, value);
}
}
return result;
},
mapValues<K, V, V2>(
map: Map<K, V>,
mapper: (value: V, key: K) => V2,
): Map<K, V2> {
const result = new Map<K, V2>();
for (const [key, value] of map) {
result.set(key, mapper(value, key));
}
return result;
},
};