saved#

Purpose#

Writes a matrix or dataframe in memory to a dataset in a specified format.

Format#

ret = saved(x, dataset[, vnames])#

Parameters:

x (NxK matrix or dataframe) – data to save
dataset (string) – name of dataset. The type of file to create is inferred from the file extension. Valid file extensions include CSV, GDAT, DAT, XLS, XLSX.
vnames (string or Kx1 string array.) –
Optional input, names for the columns of the dataset. If vnames is not passed in:
- Dataframe variable names will be used.
- Matrix data will be saved with variable names X1, X2…XP.

Returns:

ret (scalar) – 1 if successful, otherwise 0.

Note

GDAT files are the new standard GAUSS dataset format. They allow you to save dataframes with string, category and date columns.

Examples#

Save a dataframe to a CSV file#

// Load data from Stata dataset to GAUSS dataframe
fname = getGAUSSHome("examples/auto2.dta");
auto = loadd(fname, "str(make) + price + cat(foreign)");

// Print the first 5 observations of the dataframe
print auto[1:5,.];

The above code will print the first 5 observations from the auto dataframe as shown below:

         make            price          foreign
  AMC Concord        4099.0000         Domestic
    AMC Pacer        4749.0000         Domestic
   AMC Spirit        3799.0000         Domestic
Buick Century        4816.0000         Domestic
Buick Electra        7827.0000         Domestic

Now we can save this data to a CSV file with the saved() command.

// Save the data
call saved(auto, "my_auto.csv");

The first four rows of the my_auto.csv will look like this:

make,price,foreign
AMC Concord,4099,Domestic
AMC Pacer,4749,Domestic
AMC Spirit,3799,Domestic

Save a matrix to a GAUSS .dat file with default variable names#

// Create some data to save
x = rndn(100, 3);

// Create the GAUSS dataset, 'mydata.dat'
// using default variable names X1, X2 and X3
call saved(x, "mydata.dat");

Save a matrix to a GAUSS .dat file with specified variable names#

Continuing with the matrix created above, we can specify the variable names in the dataset as shown below:

// Create a 3x1 string array containing the variable names
vnames = "GDP" $| "Imports" $| "Exports";

// Create the GAUSS dataset, 'mydata.dat'
// with variable names GDP, Imports and Exports
call saved(x, "mydata.dat", vnames);

Save data to a .xlsx file#

To save the data to an Excel file, all we have to change is the file extension. Continuing with the data from the example above:

// Create an Excel dataset, 'mydata.xlsx'
call saved(x, "mydata.xlsx", vnames);

The variable names will be written as strings along the first row of the Excel file and the data will start in cell A2.

Save data to a .csv file#

To save the data to as a comma separated text file, all we have to change is the file extension. Continuing with the data from our first example:

// Create a CSV dataset, 'mydata.csv'
call saved(x, "mydata.csv", vnames);

Error checking#

The return value of saved() can be used to check whether the dataset save was successful. The example below checks the return value and creates an error if the save fails.

x = rndn(100, 2);
dataset = "mydata.dat";

// Create a 2x1 string array containing the variable names
vnames = "Price" $| "Quantity";

// Check to see if save is successful. If not, report an error and end the program
if not saved(x, dataset, vnames);
   errorlog "saved failed to write: "$+dataset;
   end;
endif;

Remarks#

You can add variable names to a matrix with dfname().

CSV

The line endings for CSV files on Windows will be \r\n and \n on Linux and macOS.
Fifteen digits of precision will be written.
csvWriteM() can be used to write CSV data with options to specify the separator to be something other than a comma, to control the line endings, or the precision to write the data.

DAT

If dataset is null or 0, the dataset name will be temp.dat.
If vnames is a null or 0, the variable names will begin with "X" and be numbered 1-K.
If vnames is a string or has fewer elements than x has columns, it will be expanded as explained under create.
The output data type is double precision.

Source#

saveload.src