Advanced data selection

The purpose of the advanced data selection is to extract larger amounts of data and download these data in file formats suitable for further processing.

The user has to select from one to 20 series, from one to 20 countries (alternatively, it is possible to select all countries), from one to 20 years and one or more "presentation type". The appropriate files are then prepared and presented to the user for download.
Note: Depending on the user's selections and on the current load on the system, the processing time may vary.

The selection process starts with the selection of series. The screen contains two list boxes: available series and selected series. Series can be selected (moved into the selected series box) by highlighting and pressing the "Select" button and can be removed from the selected series list by highlighting and pressing the "Remove" button.

The list of the available series can be filtered using the search mechanism. Two types of searches are available: "By any word" (OR type search) returns all series which contain 1) AT LEAST one of the search arguments in its name and "By all words" (AND type search), which returns those series that contain ALL of the search arguments in its name.

After a search, one can restore the full list of series by clearing the search arguments box and pressing the "Search" button again. In other words, submitting a search with no arguments, restores the full list of series.

The selections already made are preserved between searches. In other words, the user can search for "balance of payments" and select one more of the series returned. She can then submit a second search e.g. for "population" and add more series to the previously selected.

After selecting the desired series (at least one series must be selected), press the "Continue" button to proceed with the selection of countries and areas.

The selection of countries and areas is very similar to the selection of the series. One can either select some countries and press the "Continue" button, or simply press the "Select all countries and continue" button to proceed with the selection of the periods or years.

The discussion on the search for series above applies to the search for countries too.

The selection of the periods is quite straightforward. Highlight the desired periods, then press "Select" to actually select them and press "Continue" to proceed to the next page, the selection of output formats.

The selection of output formats page gives the user a consolidated view of her selections, so she can verify them and correct if necessary by backing to the appropriate selection screens. The units of presentation (millions, billions) are also selected here. The default unit is million, which is obviously inappropriate for indexes and rates. Keep also in mind that the same presentation unit is applied to all series selected. Avoid selecting series such as GDP (which is probably best presented in millions or billions) and GDP deflator, which is best viewed in units with several decimals. If you need series which require different rescale factors, extract them separately.

The output formats available are ASCII (Comma delimited values), four types of spreadsheets (MS Excel) and MS Access database. The formats are not mutually exclusive -- one can request the same data to be presented in more that one format.

The ASCII/CSV format allows for the fastest extraction. This format is probably best suitable for uploading the data in other programs such as SPSS and/or SAS and least suitable for reading (by a human being). In addition, footnotes are not included in this format. The structure of the data in such files closely resembles their internal database representation. The description of the comma separated fields is as follows:

series name;
country name;
first classification category or "n/a";
second classification category or "n/a";
third classification category or "n/a";
fourth classification category or "n/a";
fifth classification category or "n/a";
year;
Month/Quarter or "Annual";
value

An example for classification is "sex", which has categories like "male", "female" and "both sexes". Another example is "Trade commodities, SITC/Rev.2", which has more then 2000 commodities as categories.

Spreadsheet "Years on top" is a table of annual data with the individual years as column headings. Each data column (one year) is followed by a footnote column, unless there are no footnotes at all. The footnote text is at the bottom of the table. The countries and areas selected are the rows of the table. Any classification associated with the series (such as male/female, urban/rural etc.) is presented in separate columns of the table. Multiple series are presented on different sheets of the spreadsheet. A maximum of 20 years can be presented.

Spreadsheet "Months on top" is a table with one column of yearly data and 12 columns of monthly data for the selected year. If more than one year is selected, only the latest one is presented in the table. Spreadsheet "Quarters on top" is very similar.

Spreadsheet "Series on top" is a table where the series are columns headings. Countries and areas and years are presented on the left-hand side of the table as separate columns. Only series that have the same classifications can be presented in a meaningful way in such a table. This presentation format may therefore not be available, depending on the selected series.

The data can also be extracted in a MS Access database format. The internal structure of the Access database closely follows the structure of the Common Database itself. While this format is suitable for extracting large volumes of data, especially if the intention is to process programmatically these data, its use may require some familiarity with databases and possibly some programming skills.

To start the preparation of the requested data, press the "Prepare data " button on the "Output format selection screen". On the next screen, "Compile requested tables", the user can check from time to time if the tables are ready. If so, appropriate links pointing to the files are displayed. Clicking on those links "downloads" or "opens" the file, depending on the settings of the browser used. In any case, the files can be saved locally on the computer of the user for later viewing or processing.
Alternatively, one can request to be notified by e-mail upon completion of the extraction. The e-mail message will contain links for downloading the data. Users are encouraged to use the e-mail notification feature.
The screen also shows the queue of requests in the vicinity of the specific request.

1) The search mechanism work on an index of all the words of the short series name, all the words of the full series name (which may contain additional keywords) and all the definitions. If the search for the words as typed fails to return any results (possibly because of misspelling), a soundex search is automatically activated. The SOUNDEX technology groups together "similarly sounding" words. On the plus side, the user does not have worry about the correct spelling. Some of the associations however may be quite surprising.