MATLAB Examples

This file is designed to help you oonvert JSON strings into MATLAB-usable data. Note that it works with a string, not a file, so you have to load in the file yourself.

## Starting with a string

Let's take a simple structure in JSON and load it. The JSON.m file is a MATLAB class with a single public method: JSON.parse()

str = '{"name":"North Carolina","capital":"Raleigh","population":"9,752,073"}';
data = JSON.parse(str)
data =

name: 'North Carolina'
capital: 'Raleigh'
population: '9,752,073'

Now we can work with the data in MATLAB.

fprintf('The capital of %s is %s\n', ...
data.name,data.capital)
The capital of North Carolina is Raleigh

## Dealing with Numbers

Be aware the numbers may be represented as strings or as numbers, depending on how they appear in the JSON text.

fprintf('The population of %s is %s\n', ...
data.name, ...
data.population)
The population of North Carolina is 9,752,073

That looks good. But if you try to do math on the population, you're in for a surprise. Suppose we try to round to the nearest thousand.

round(data.population/1000)*1000
ans =

0     0     0     0     0     0     0     0     0

Not exactly what we want. So we can use str2double like so:

round(str2double(data.population)/1000)*1000
ans =

9752000

If we are the ones making the JSON string in the first place, it might be preferable to store the number without the quotes. In this case

str = '{"name":"North Carolina","capital":"Raleigh","population":9752073}';
data = JSON.parse(str)

fprintf('Rounded to the nearest thousand, the population of %s is %d\n', ...
data.name, ...
round(data.population/1000)*1000)
data =

name: 'North Carolina'
capital: 'Raleigh'
population: 9752073

Rounded to the nearest thousand, the population of North Carolina is 9752000

Generally your JSON text will come from a file.

fname = 'capitals.json';
fid = fopen(fname);
str = char(raw');
fclose(fid);

data = JSON.parse(str)
data =

Columns 1 through 4

[1x1 struct]    [1x1 struct]    [1x1 struct]    [1x1 struct]

Column 5

[1x1 struct]

The result in this case is a cell array of structures.

for n = 1:length(data)
fprintf('The capital of %s is %s\n', ...
data{n}.name,data{n}.capital)
end
The capital of Alabama is Montgomery
The capital of Georgia is Atlanta
The capital of New York is Albany
The capital of Missouri is Jefferson City
The capital of North Carolina is Raleigh