ANY-maze Help > The ANY-maze reference > The Data page > Transferring data to specific programs > Transferring data to Statistica

Transferring data to Statistica

Introduction

Statistica is a popular statistical analysis program. Transferring data from ANY-maze to Statistica is a little complex, as all the possible methods involve some degree of compromise.

If your data includes repeated measures, you should use the options to Show repeated measures in columns, as this is the format that Statistica expects.

Copy and paste

Copying data is probably the best way to transfer data to Statistica. However, you should note that if you copy the entire ANY-maze spreadsheet (i.e. without selecting anything), then when you paste it into Statistica, the column titles will appear as the first case. You can avoid this by either deleting this case or by selecting all the rows in the Data page spreadsheet before copying them - see Selecting cells, columns or rows for details.

Another thing to be aware of is that by default, Statistica creates New data tables with just 10 variables and 10 cases (i.e. 10 columns and 10 rows). If you paste more data than this into a new table, the excess columns and/or rows are simply ignored. To overcome this, you should add enough variables and cases to accommodate all your data before you paste. In fact, adding too many rows or columns isn't a problem, so it's often easiest to simply add, say, 100 variables and 1000 cases.

Saving data as comma (CSV) or tab (TXT) separated files

Statistica can read comma (CSV) or tab (TXT) separated text files, so saving data in either format and then opening the file in Statistica is another way to transfer data.

However, this doesn't work as well as you might expect, because Statistica truncates any text data to 8 characters. For example, if you have two treatments called 'Compound ABC 1mg/kg' and 'Compound ABC 2mg/kg', they will both be chopped off at 'Compound' - in fact Statistica will call one of them 'Compound' and the other 'Compoun2', but knowing which is which is still rather difficult. Nevertheless, if you plan read data into Statistica then you can just take care to use short names.

Another more serious issue is that variable names (which are taken from column titles) are also truncated - this can lead to some very confusing names, and for this reason using dBase format files is better.

Saving data in a dBase file

Saving a Data page spreadsheet in dBase format and then reading it into Statistica is better than using text format, because ANY-maze will title the dBase fields with 8 character short 'variable names' which will help you differentiate the variables. However, despite the fact that the dBase format specifies column widths, Statistica will still insist on truncating strings to 8 characters, as it does when reading text files - again, keeping names short is the only viable solution to this problem.

How to open a CSV, TXT or dBase file in Statistica

These instructions are for Statistica version 6.0

Click the File menu.

Select Import Data | Quick...; the File import window will appear.

Towards the bottom of the window is a list titled Files of Type. Select either 'Text' or 'dBase', depending on the type of file you wish to open. (Note: for CSV files, select 'Text').

In the main part of the window, navigate to the file you saved from within ANY-maze and select it.

Click the Open button.

When you open a text file, the Quick import from text window will appear. If the file is a CSV file, then select Comma as the Field separator; otherwise select Tab. At the bottom of the window, select the option to Get variable names from the first row.