readr#

Purpose#

Reads a specified number of rows of data from a GAUSS dataset (.dat) file, a GAUSS matrix (.fmt) file, or an HDF5 (.h5) file.

Format#

y = readr(f1, r)#

Parameters:

f1 (scalar) – file handle of an open file.
r (scalar) – number of rows to read.

Returns:

y (NxK matrix) – the data read from the file.

Examples#

Basic example#

// Get file name with full path
fname = getGAUSSHome("examples/cancer.dat");

// Open file handle for reading only
fh = dataOpen(fname, "read");

// Read the first 2 rows from the dataset
x = readr(fh, 2);

// Close file handle
call close(fh);

After the above code, x will equal:

1    1    1    9  157
1    2    1    5   77

Iteratively read an entire dataset#

// Get file name with full path
fname = getGAUSSHome("examples/cancer.dat");

// Open file handle for reading only
fh = dataOpen(fname, "read");

sum = 0;

// Continue the loop until the
// end of file is found
do until eof(fh);
   // Read 20 rows per iteration
   x = readr(fh, 20);
   sum = sum + sumc(x);
endo;

call close(fh);

This code reads 20 rows from a dataset at a time. The sum of the columns is computed and added to the previous sum.

The result is the sum of the columns for the entire dataset. eof(fh) returns 1 when the end of the dataset is encountered.

Remarks#

The first time a readr() statement is encountered, the first r rows will be read. The next time it is encountered, the next r rows will be read in, and so on. If the end of the dataset is reached before r rows can be read, then only those rows remaining will be read.

After the last row has been read, the pointer is placed immediately after the end of the file. An attempt to read the file in these circumstances will cause an error message.

To move the pointer to a specific place in the file use seekr().