Skip to main content

Variable: boxplot

const boxplot: ExtendedTool<BoxplotToolArgs, BoxplotLlmResult, BoxplotAdditionalData, EChartsToolContext>

Defined in: packages/echarts/src/boxplot/tool.ts:49

The boxplot tool is used to create a box plot for a given dataset and variable.

Example​

import { getVercelAiTool } from '@openassistant/plots';
import { generateText } from 'ai';

const toolContext = {
  getValues: async (datasetName: string, variableName: string) => {
    return SAMPLE_DATASETS[datasetName].map((item) => item[variableName]);
  },
};

const onToolCompleted = (toolCallId: string, additionalData?: unknown) => {
  console.log('Tool call completed:', toolCallId, additionalData);
  // render the boxplot using <BoxplotComponentContainer props={additionalData} />
};

const boxplotTool = getVercelAiTool('boxplot', toolContext, onToolCompleted);

generateText({
  model: openai('gpt-4o-mini', { apiKey: key }),
  prompt: 'Can you create a box plot of the revenue per capita for each location in dataset myVenues?',
  tools: {boxplot: boxplotTool},
});

getValues()​

See BoxplotFunctionContext for detailed usage.

User implements this function to get the values of the variable from dataset.

For prompts like "can you show a box plot of the revenue per capita for each location in dataset myVenues", the tool will call the getValues() function twice:

  • get the values of revenue from dataset: getValues('myVenues', 'revenue')
  • get the values of population from dataset: getValues('myVenues', 'population')

A duckdb table will be created using the values returned from getValues(), and LLM will generate a sql query to query the table to answer the user's prompt.