read

Read data in datastore

Syntax

data = read(ds)

[data,info] = read(ds)

Description

data = read(ds) returns data from a datastore. Subsequent calls to the read function continue reading from the endpoint of the previous call.

[data,info] = read(ds) also returns information about the extracted data in info, including metadata.

Examples

Read Data in TabularTextDatastore

Create a datastore from the sample file, airlinesmall_subset.csv, which contains tabular data.

ds = tabularTextDatastore("airlinesmall_subset.csv","TreatAsMissing","NA","MissingValue",0);

Modify the SelectedVariableNames property to specify the variables of interest.

ds.SelectedVariableNames = ["DepTime","ArrTime","ActualElapsedTime"];

While there is data available to be read from the datastore, read one block of data at a time and analyze the data. In this example, sum the actual elapsed time.

sumElapsedTime = 0;
while hasdata(ds)
    T = read(ds);
    sumElapsedTime = sumElapsedTime + sum(T.ActualElapsedTime);
end

View the sum of the actual elapsed time.

sumElapsedTime

sumElapsedTime = 
154759

Read Data in KeyValueDatastore

Create a datastore from the sample file, mapredout.mat, which is the output file of the mapreduce function.

ds = datastore('mapredout.mat');

Read a subset of data in the datastore.

T = read(ds)

T=1×2 table
     Key        Value  
    ______    _________

    {'AA'}    {[14930]}

Change the number of key-value pairs to read at a time, by changing the ReadSize property of the datastore.

ds.ReadSize = 5;

Read the next five key-value pairs in the datastore.

T = read(ds)

T=5×2 table
     Key        Value  
    ______    _________

    {'AS'}    {[ 2910]}
    {'CO'}    {[ 8138]}
    {'DL'}    {[16578]}
    {'EA'}    {[  920]}
    {'HP'}    {[ 3660]}

Read Data in CombinedDatastore

Create a datastore that maintains parity between the pair of images of the underlying datastores. For instance, create two separate image datastores, and then create a combined datastore that reads corresponding images from the two image datastores.

Create an image datastore imds1 representing a collection of three images.

imds1 = imageDatastore({'peppers.png','street1.jpg','street2.jpg'}); 

Create a second datastore imds2 containing a mask of the bright regions of the three images. To create this datastore, first transform the images of imds1 to grayscale. Then convert each image to a binary mask by performing thresholding. In this example, the thresholding operation maps pixels with a value above the threshold (250) to white and all other pixels to black.

imds2 = transform(imds1,@(x) im2gray(x)>250);

Create a combined datastore from imds1 and imds2.

imdsCombined = combine(imds1,imds2);

Read the first subset of data from the combined datastore. The output is a 1-by-2 cell array. The two columns represent the first subset of data read from the two underlying datastores imds1 and imds2, respectively.

dataOut = read(imdsCombined)

dataOut=1×2 cell array
    {384x512x3 uint8}    {384x512 logical}

Display the read data from the combined datastore as a pair of tiled images.

tile = imtile(dataOut);
imshow(tile)

Read from the combined datastore again. This call to the read function continues reading from the endpoint of the previous call.

dataOut = read(imdsCombined)

dataOut=1×2 cell array
    {480x640x3 uint8}    {480x640 logical}

Display the read data.

tile = imtile(dataOut);
imshow(tile)

Input Arguments

ds — Input datastore

datastore

Input datastore. You can use these datastores as input to the read method.

• MATLAB^® datastores — Datastores created using MATLAB datastore functions. For example, create a datastore for a collection of images using ImageDatastore. For a complete list of datastores, see Select Datastore for File Format or Application.

• Combined and transformed datastores — Datastores created using the combine and transform functions.

• Custom datastores — Datastores created using the custom datastore framework. See Develop Custom Datastore.

Output Arguments

data — Output data

table | timetable | array

Output data, returned as a table, timetable, or array depending on the type of the input ds. The data output can be empty if the amount of data read in a particular call to the read function in combination with the import configuration returns no values.

Type of Datastore	Data type of data	Description
TabularTextDatastore, SpreadsheetDatastore, and ParquetDatastore	Table or timetable	The SelectedVariableNames property determines the table variables. The OutputType property determines if the output is a table or timetable.
ImageDatastore	Integer array	The dimensions of the integer array depend on the type of image: For grayscale images, data is m-by-n. For truecolor images, data is m-by-n-by-3. For CMYK Tiff images, data is m-by-n-by-4. If the ReadSize property is greater than 1, then data is a cell array of image data corresponding to each image. The read function supports all image types supported by the imread function. For more information on the supported image types, see imread.
KeyValueDatastore	Table	The table variable names are Key and Value.
FileDatastore	Varies	The output is the same as the output returned by the custom read function, specified by the 'ReadFcn' value.
TransformedDatastore	Varies	The output is the same as the output of the transformation function @fcn specified in the transform method used to create the TransformedDatastore.
CombinedDatastore	Varies	Contains the horizontal concatenation of the output of read from the corresponding underlying datastores.
SequentialDatastore	Varies	Contains the output of sequential read from the current underlying datastore.

info — Information about read data

structure array | cell array

Information about read data, returned as a structure array or a cell array of structure arrays.

• For MATLAB datastores and TransformedDatastore, info is a structure array that has fields with information about the datastore.

• For CombinedDatastore, info is a cell array of structure arrays. Each element of the cell array contains a structure with the relevant fields of the corresponding underlying datastore.

• For SequentialDatastore, the data type and format of info are the same as the current underlying datastore.

Information in the structure array depends on the type of the input datastore. The structure array can contain the following fields.

Field Name	Datastore Types	Description
Filename	ImageDatastore,SpreadsheetDatastore, TabularTextDatastore, FileDatastore, KeyValueDatastore, and TallDatastore	Filename is a fully resolved path containing the path string, name of the file, and file extension. For ImageDatastore objects whose ReadSize property is greater than 1, Filename is a cell array of file names corresponding to each image.
FileSize		Total file size, in bytes. For ImageDatastore objects whose ReadSize property is greater than 1, FileSize is a vector of file sizes corresponding to each image. For MAT-files, the value of FileSize depends on the type of the datastore. KeyValueDatastore and TallDatastore — The FileSize field contains the total number of key-value pairs in the file. FileDatastore — The FileSize field contains the total file size in bytes.
FileType	KeyValueDatastore only	The type of file from which data is read, either 'mat' for MAT-files or 'seq' for sequence files.
Label	ImageDatastore only	Image label name. If the ReadSize property is greater than 1, then Label is a vector of label names corresponding to each image. If the Labels property is empty, then Label is an empty cell array.
NumCharactersRead	TabularTextDatastore only	Number of characters read.
NumDataRows	SpreadsheetDatastore only	Vector containing number of rows read from each sheet.
Offset	KeyValueDatastore and TabularTextDatastore only	Starting position of the read operation, in bytes. For MAT-files, Offset is the index of the first key and value read.
SheetNames	SpreadsheetDatastore only	Names of sheets read.
SheetNumbers	SpreadsheetDatastore only	Numbering associated with sheets read.

Extended Capabilities

Thread-Based Environment

Run code in the background using MATLAB® backgroundPool or accelerate code with Parallel Computing Toolbox™ ThreadPool.

Usage notes and limitations:

• In a thread-based environment, you can use read only with the following datastores:

  • ImageDatastore objects

  • CombinedDatastore, SequentialDatastore, or TransformedDatastore objects you create from ImageDatastore objects by using combine or transform

  You can use read with other datastores if you have Parallel Computing Toolbox™. To do so, run the function using a process-backed parallel pool instead of using backgroundPool or ThreadPool (use either ProcessPool or ClusterPool).

• read does not support the following image formats in a thread-based environment:

  • Hierarchical Data Format (HDF)

  • SVS

  • TIFF

For more information, see Run MATLAB Functions in Thread-Based Environment.

Version History

Introduced in R2014b