JSON data basics with Tablesaw

Importing Data in Java

Anthony Markham

VP Quant Developer

JSON introduction

JSON: JavaScript Object Notation
- Lightweight data interchange format
- Web API standard format
- Key-value pair structure

Tabular data

Fixed rows and columns
Tabular: Uniform data types

| Name  | Age | City    |
|-------|-----|---------|
| Alice | 30  | Boston  |
| Bob   | 25  | Seattle |

JSON

Flexible nested structures
Mixed data types and hierarchies

[
  {"name": "Alice", 
      "age": 30, 
      "address": {"city": "Boston", 
                  "state": "MA"}},
  {"name": "Bob", 
      "age": 25, 
      "address": {"city": "Seattle", 
                  "state": "WA"}}
]

Reading JSON

Use only for simple JSON files

// Simple method to read a JSON file
Table products = Table.read().file("products.json");

import tech.tablesaw.io.json.JsonReader;
import tech.tablesaw.io.json.JsonReadOptions;

Offers additional configuration

// Reading a JSON file using JsonReadOptions
JsonReadOptions options = JsonReadOptions.builder("products.json").build();

Table products = new JsonReader().read(options);

Accessing JSON data

Familiar Tablesaw methods work

// Access columns from JSON data
StringColumn names = table.stringColumn("name");
DoubleColumn prices = table.doubleColumn("price");


// Perform calculations
double avgPrice = prices.mean();
String mostExpensive = table
    .where(prices.isEqualTo(prices.max()))
    .stringColumn("name").get(0);

Best JSON practices - validation

Validate JSON structure before processing

// Validation and error handling
try {
    JsonReadOptions options = JsonReadOptions.builder("data.json").build();
    Table data = new JsonReader().read(options);
    if (data.rowCount() > 0) {
        // Process data
    }
} catch (Exception e) {
    System.err.println("Error reading JSON: " + e.getMessage());
}

if (data.rowCount() == 100) {
    System.out.println("Table has exactly 100 rows - processing data");
}

Best JSON practices - missing values

Handle missing/null values

// Remove rows with any missing values
data = data.dropRowsWithMissingValues();

Match data types to analysis needs

// Convert integer column to double for calculations
data = data.replaceColumn("price", data.intColumn("price").asDoubleColumn());

Let's practice!

Importing Data in Java