Load

The LOAD statement loads fields from a file, from data defined in the script, from a previously loaded table, from a web page, from the result of a subsequent SELECT statement or by generating data automatically. It is also possible to load data from analytic connections.

Syntax:

LOAD [ distinct ] fieldlist

[( from file [ format-spec ] |

from_field fieldassource [format-spec]|

inline data [ format-spec ] |

resident table-label |

autogenerate size ) |extension pluginname.functionname([script] tabledescription)]

[ where criterion | while criterion ]

[ group by groupbyfieldlist ]

[order by orderbyfieldlist ]

Arguments:

Arguments
Argument	Description
distinct	You can use distinct as a predicate if you only want to load unique records. If there are duplicate records, the first instance will be loaded. If you are using preceding loads, you need to place distinct in the first load statement, as distinct only affects the destination table. Preceding load
fieldlist	fieldlist ::= ( * \| field{, * \| field } ) A list of the fields to be loaded. Using * as a field list indicates all fields in the table. field ::= ( fieldref \| expression ) [as aliasname ] The field definition must always contain a literal, a reference to an existing field, or an expression. fieldref ::= ( fieldname \|@fieldnumber \|@startpos:endpos [ I \| U \| R \| B \| T] ) fieldname is a text that is identical to a field name in the table. Note that the field name must be enclosed by straight double quotation marks or square brackets if it contains e.g. spaces. Sometimes field names are not explicitly available. Then a different notation is used: @fieldnumber represents the field number in a delimited table file. It must be a positive integer preceded by "@". The numbering is always made from 1 and up to the number of fields. @startpos:endpos represents the start and end positions of a field in a file with fixed length records. The positions must both be positive integers. The two numbers must be preceded by "@" and separated by a colon. The numbering is always made from 1 and up to the number of positions. In the last field, n is used as end position. If @startpos:endpos is immediately followed by the characters I or U, the bytes read will be interpreted as a binary signed (I) or unsigned (U) integer (Intel byte order). The number of positions read must be 1, 2 or 4. If @startpos:endpos is immediately followed by the character R, the bytes read will be interpreted as a binary real number (IEEE 32-bit or 64 bit floating point). The number of positions read must be 4 or 8. If @startpos:endpos is immediately followed by the character B, the bytes read will be interpreted as a BCD (Binary Coded Decimal) numbers according to the COMP-3 standard. Any number of bytes may be specified. expression can be a numeric function or a string function based on one or several other fields in the same table. For further information, see the syntax of expressions. as is used for assigning a new name to the field.
from	from is used if data should be loaded from a file using a folder or a web file data connection file ::= [ path ] filename Example: 'lib://Table Files/' If the path is omitted, Qlik Sense searches for the file in the directory specified by the Directory statement. If there is no Directory statement, Qlik Sense searches in the working directory, C:\Users\{user}\Documents\Qlik\Sense\Apps. Directory Information noteIn a Qlik Sense server installation, the working directory is specified in Qlik Sense Repository Service, by default it is C:\ProgramData\Qlik\Sense\Apps. See the Qlik Management Console help for more information. The filename may contain the standard DOS wildcard characters ( * and ? ). This will cause all the matching files in the specified directory to be loaded. format-spec ::= ( fspec-item { , fspec-item } ) The format specification consists of a list of several format specification items, within brackets. Format specification items Tip noteYou can use the URL is format specification to override the URL of a web file data connection, for example if you need to create a dynamic URL based on other data that was loaded. Legacy scripting mode In legacy scripting mode, the following path formats are also supported: absolute Example: c:\data\ relative to the Qlik Sense app working directory. Example: data\ URL address (HTTP or FTP), pointing to a location on the Internet or an intranet. Example: http://www.qlik.com
from_field	from_field is used if data should be loaded from a previously loaded field. fieldassource::=(tablename, fieldname) The field is the name of the previously loaded tablename and fieldname. format-spec ::= ( fspec-item {, fspec-item } ) The format specification consists of a list of several format specification items, within brackets. For more information, see Format specification items. Information notefrom_field only supports commas as the list delimiter in when separating fields in tables.
inline	inline is used if data should be typed within the script, and not loaded from a file. data ::= [ text ] Data entered through an inline clause must be enclosed by double quotation marks or by square brackets. The text between these is interpreted in the same way as the content of a file. Hence, where you would insert a new line in a text file, you should also do it in the text of an inline clause, i.e. by pressing the Enter key when typing the script. The number of columns are defined by the first line. format-spec ::= ( fspec-item {, fspec-item } ) The format specification consists of a list of several format specification items, within brackets. For more information, see Format specification items.
resident	resident is used if data should be loaded from a previously loaded table. table label is a label preceding the LOAD or SELECT statement(s) that created the original table. The label should be given with a colon at the end. Loading data from a previously loaded table Table labels
autogenerate	autogenerate is used if data should be automatically generated by Qlik Sense. size ::= number Number is an integer indicating the number of records to be generated. The field list must not contain expressions which require data from an external data source or a previously loaded table, unless you refer to a single field value in a previously loaded table with the Peek function.
extension	You can load data from analytic connections. You need to use the extension clause to call a function defined in the server-side extension (SSE) plugin, or evaluate a script. You can send a single table to the SSE plugin, and a single data table is returned. If the plugin does not specify the names of the fields that are returned, the fields will be named Field1, Field2, and so on. Extension pluginname.functionname( tabledescription ); Loading data using a function in an SSE plugin tabledescription ::= (table { ,tablefield} ) If you do not state table fields, the fields will be used in load order. Loading data by evaluating a script in an SSE plugin tabledescription ::= ( script, table { ,tablefield} ) Data type handling in the table field definition Data types are automatically detected in analytic connections. If the data has no numeric values and at least one non-NULL text string, the field is considered as text. In any other case it is considered as numeric. You can force the data type by wrapping a field name with String() or Mixed(). String() forces the field to be text. If the field is numeric, the text part of the dual value is extracted, there is no conversion performed. Mixed() forces the field to be dual. String() or Mixed() cannot be used outside extension table field definitions, and you cannot use other Qlik Sense functions in a table field definition. More about analytic connections You need to configure analytic connections before you can use them. Qlik Sense Enterprise: Creating an analytic connection Qlik Sense Desktop: Configuring analytic connections in Qlik Sense Desktop You can read more about analytic connections in the GitHub repository. qlik-oss/server-side-extension
where	where is a clause used for stating whether a record should be included in the selection or not. The selection is included if criterion is True. criterion is a logical expression.
while	while is a clause used for stating whether a record should be repeatedly read. The same record is read as long as criterion is True. In order to be useful, a while clause must typically include the IterNo( ) function. criterion is a logical expression.
group by	group by is a clause used for defining over which fields the data should be aggregated (grouped). The aggregation fields should be included in some way in the expressions loaded. No other fields than the aggregation fields may be used outside aggregation functions in the loaded expressions. groupbyfieldlist ::= (fieldname { ,fieldname } )
order by	order by is a clause used for sorting the records of a resident table before they are processed by the load statement. The resident table can be sorted by one or more fields in ascending or descending order. The sorting is made primarily by numeric value and secondarily by national collation order. This clause may only be used when the data source is a resident table. The ordering fields specify which field the resident table is sorted by. The field can be specified by its name or by its number in the resident table (the first field is number 1). orderbyfieldlist ::= fieldname [ sortorder ] { , fieldname [ sortorder ] } sortorder is either asc for ascending or desc for descending. If no sortorder is specified, asc is assumed. fieldname, path, filename and aliasname are text strings representing what the respective names imply. Any field in the source table can be used as fieldname. However, fields created through the as clause (aliasname) are out of scope and cannot be used inside the same load statement.

If no source of data is given by means of a from, inline, resident, from_field, extension or autogenerate clause, data will be loaded from the result of the immediately succeeding SELECT or LOAD statement. The succeeding statement should not have a prefix.

Loading data from a previously loaded table

Examples:

Load a delimited data file with default options:

LOAD * from data1.csv;

Load a delimited data file from a library connection (DataFiles):

LOAD * from 'lib://DataFiles/data1.csv';

Load all delimited data files from a library connection (DataFiles):

LOAD * from 'lib://DataFiles/*.csv';

Load a delimited file, specifying comma as delimiter and with embedded labels:

LOAD * from 'c:\userfiles\data1.csv' (ansi, txt, delimiter is ',', embedded labels);

Load a delimited file specifying tab as delimiter and with embedded labels:

LOAD * from 'c:\userfiles\data2.txt' (ansi, txt, delimiter is '\t', embedded labels);

Load a dif file with embedded headers:

LOAD * from file2.dif (ansi, dif, embedded labels);

Load three fields from a fixed record file without headers:

LOAD @1:2 as ID, @3:25 as Name, @57:80 as City from data4.fix (ansi, fix, no labels, header is 0, record is 80);

Load a QVX file, specifying an absolute path:

LOAD * from C:\qdssamples\xyz.qvx (qvx);

Load from the default URL set in the web file data connection:

LOAD * from [lib://MyWebFile];

Load from a specific URL, and override the URL set in the web file data connection:

LOAD * from [lib://MyWebFile] (URL is 'http://localhost:8000/foo.bar');

Load from a specific URL set in a variable using dollar-sign expansion:

SET dynamicURL = 'http://localhost/foo.bar';

LOAD * from [lib://MyWebFile] (URL is '$(dynamicURL)');

Load only three specific fields from a delimited file:

LOAD FirstName, LastName, Number from data1.csv;

Rename first field as A and second field as B when loading a file without labels:

LOAD @1 as A, @2 as B from data3.txt (ansi, txt, delimiter is '\t', no labels);

Load Name as a concatenation of FirstName, a space character, and LastName:

LOAD FirstName&' '&LastName as Name from data1.csv;

Load Quantity, Price and Value (the product of Quantity and Price):

LOAD Quantity, Price, Quantity*Price as Value from data1.csv;

Load only unique records, duplicate records will be discarded:

LOAD distinct FirstName, LastName, Number from data1.csv;

Load only records where the field Litres has a value above zero:

LOAD * from Consumption.csv where Litres>0;

Load a table with inline data, two fields named CatID and Category:

LOAD * Inline

[CatID, Category

0,Regular

1,Occasional

2,Permanent];

Load a table with inline data, three fields named UserID, Password and Access:

LOAD * Inline [UserID, Password, Access

A, ABC456, User

B, VIP789, Admin];

Load a table with 10 000 rows. Field A will contain the number of the read record (1,2,3,4,5...) and field B will contain a random number between 0 and 1:

LOAD RecNo( ) as A, rand( ) as B autogenerate(10000);

The parenthesis after autogenerate is allowed but not required.

First we load a delimited table file and name it tab1:

tab1:

SELECT A,B,C,D from 'lib://DataFiles/data1.csv';

Load fields from the already loaded tab1 table as tab2:

tab2:

LOAD A,B,month(C),A*B+D as E resident tab1;

Load fields from already loaded table tab1 but only records where A is larger than B:

tab3:

LOAD A,A+B+C resident tab1 where A>B;

Load fields from already loaded table tab1 ordered by A:

LOAD A,B*C as E resident tab1 order by A;

Load fields from already loaded table tab1, ordered by the first field, then the second field:

LOAD A,B*C as E resident tab1 order by 1,2;

Load fields from already loaded table tab1 ordered by C descending, then B in ascending order, and then the first field in descending order:

LOAD A,B*C as E resident tab1 order by C desc, B asc, 1 desc;

Load field Types from previously loaded table Characters as A:

LOAD A from_field (Characters, Types);

Load A, B and calculated fields X and Y from Table1 that is loaded in succeeding SELECT statement:

LOAD A, B, if(C>0,'positive','negative') as X, weekday(D) as Y;

SELECT A,B,C,D from Table1;

Load fields grouped (aggregated) by ArtNo:

LOAD ArtNo, round(Sum(TransAmount),0.05) as ArtNoTotal from table.csv group by ArtNo;

Load fields grouped (aggregated) by Week and ArtNo:

LOAD Week, ArtNo, round(Avg(TransAmount),0.05) as WeekArtNoAverages from table.csv group by Week, ArtNo;

In this example we have a input file Grades.csv containing the grades for each student condensed in one field:

Student,Grades

Mike,5234

John,3345

Pete,1234

Paul,3352

The grades, in a 1-5 scale, represent subjects Math, English, Science and History. We can separate the grades into separate values by reading each record several times with a while clause, using the IterNo( ) function as a counter. In each read, the grade is extracted with the Mid function and stored in Grade, and the subject is selected using the pick function and stored in Subject. The final while clause contains the test to check if all grades have been read (four per student in this case), which means next student record should be read.

MyTab:

LOAD Student,

mid(Grades,IterNo( ),1) as Grade,

pick(IterNo( ), 'Math', 'English', 'Science', 'History') as Subject from Grades.csv

while IsNum(mid(Grades,IterNo(),1));

The result is a table containing this data:

Resulting table data

The following sample data is used.

Values: Load Rand() as A, Rand() as B, Rand() as C AutoGenerate(50);

Loading data using a function

In these examples, we assume that we have an analytic connection plugin named P that contains a custom function Calculate(Parameter1, Parameter2). The function returns the table Results that contains the fields Field1 and Field2.

Load * Extension P.Calculate( Values{A, C} );

Load all fields that are returned when sending the fields A and C to the function.

Load Field1 Extension P.Calculate( Values{A, C} );

Load only the Field1 field when sending the fields A and C to the function.

Load * Extension P.Calculate( Values );

Load all fields that are returned when sending the fields A and B to the function. As fields are not specified, A and B are used as they are the first in order in the table.

Load * Extension P.Calculate( Values {C, C});

Load all fields that are returned when sending the field C to both parameters of the function.

Load * Extension P.Calculate( Values {String(A), Mixed(B)});

Load all fields that are returned when sending the field A forced as a string and B forced as a numeric to the function.

Loading data by evaluating a script

Load A as A_echo, B as B_echo Extension R.ScriptEval( 'q;', Values{A, B} );

Load the table returned by the script q when sending the values of A and B.

Load * Extension R.ScriptEval( '$(My_R_Script)', Values{A, B} );

Load the table returned by the script stored in the My_R_Script variable when sending the values of A and B.

Load * Extension R.ScriptEval( '$(My_R_Script)', Values{B as D, *} );

Load the table returned by the script stored in the My_R_Script variable when sending the values of B renamed to D, A and C. Using * sends the remaining unreferenced fields.

The file extension of DataFiles connections is case sensitive. For example: .qvd.

Did this page help you?

If you find any issues with this page or its content – a typo, a missing step, or a technical error – let us know how we can improve!

Leave your feedback here