Adrenaline DataFlow Functions Reference Guide

IFNULL

Return the first parameter if non null, otherwise return the second parameter

IFNULL(ANY, ANY) -> ANY

NULLIF

Return null if the two input expressions are equal, otherwise return the first expression.

NULLIF(ANY, ANY) -> ANY

NVL

Similar to COALESCE but only accepts exactly two input expressions

NVL(ANY, ANY) -> ANY

CEILING

Takes in a number and returns the nearest whole number rounded up.

CEILING(NUMERIC) -> LONG

FLOOR

Takes in a number and returns ths nearest whole number rounded down.

FLOOR(NUMERIC) -> LONG

LN

Return the natural log of the input parameter.

LN(NUMERIC) -> DOUBLE

LOG

Return the base 10 log of the input parameter.

LOG(NUMERIC) -> DOUBLE

POWER

Returns the first number raised to the power of the second number.

POWER(NUMERIC, NUMERIC) -> DOUBLE

RAND

Returns a uniformly-distributed random number x, where 0 <= x < 1. Takes a parameter of any type as a seed.

RAND(ANY) -> DOUBLE

RANDOM

Returns a uniformly-distributed random number x, where 0 <= x < 1.

RANDOM() -> DOUBLE

ROUND

Rounds a value to a specified number of decimal places, retaining the original precision and scale. Fractions greater than or equal to .5 are rounded up. Fractions less than .5 are rounded down (truncated).

ROUND(NUMERIC) -> DOUBLE

SQRT

Returns the square root of the input parameter. Returns null if the result is not a real number.

SQRT(NUMERIC) -> DOUBLE

WIDTH_BUCKET

Constructs equiwidth histograms, in which the histogram range is divided into intervals (buckets) of identical sizes. In addition, values below the low bucket return 0, and values above the high bucket return bucket_count +1. Returns an integer value.

WIDTH_BUCKET(NUMERIC|TIME expression, NUMERIC|TIME min, NUMERIC|TIME max, LONG bucket_count) -> LONG

ADDTIME

Adds a time value to a date or datetime. The time is represented as the timestamp component of a string date time value.

ADDTIME(TIME, TIME_LITERAL) -> TIME

CONVERT_TZ

Converts a datetime value from one timezone to another. Timezones can be hour offsets, timezone codes, or full timezone names (e.g. -06:00, MDT, America/Denver)

CONVERT_TZ(DATETIME, STRING from_tz, STRING to_tz) -> DATETIME

DATE

Extracts the date portion from a datetime expression.

DATE(TIME) -> DATE

DATE_TRUNC

Truncates a datetime expression to part specified.

DATE_TRUNC(STRING date_part, TIME) -> TIME

DATEDIFF

Returns the number of days between two date values

DATEDIFF(TIME, TIME) -> LONG

DAYNAME

Return the full name of a the weekday for a date.

DAYNAME(TIME) -> STRING

DAYOFMONTH

Return the day of the month for a date.

DAY(TIME) -> LONG

DAYOFWEEK

Return the day number in the week for a date. The return value is the same as WEEKDAY.

DAYOFWEEK(TIME) -> LONG

DAYOFYEAR

Return the day number in the year for a date.

DAYOFYEAR(TIME) -> LONG

FROM_DAYS

Returns a date from a numerical representation of the date (days since epoch). See TO_DAYS.

FROM_DAYS(NUMERIC) -> DATE

FROM_UNIXTIME

Returns a date from a unix based time (seconds since epoch). Optionally a pattern for the output format can be included. See UNIX_TIMESTAMP.

FROM_UNIXTIME(NUMERIC, [STRING pattern]) -> DATETIME

HOUR

Returns the hour portion of a datetime value.

HOUR(TIME) -> LONG

LAST_DAY

Returns the last day of the month for a given date or datetime.

LAST_DAY(TIME) -> DATE

MONTH

Returns the month number for a given date or datetime.

MONTH(TIME) -> LONG

MONTHNAME

Returns the full name of the month for a given date or datetime.

MONTHNAME(TIME) -> STRING

NOW

Returns the current time in the timezone of the query request context.

NOW() -> DATETIME

PERIOD_ADD

Adds a number of months to a period. Period is a number formatted YYYYMM or YYMM. The result is a long formatted as YYYYMM.

PERIOD_ADD(NUMERIC period, NUMERIC months) -> LONG

PERIOD_DIFF

Difference in months between two periods. Period is a number formatted YYYYMM or YYMM. Both periods should be the same format. The result is the number of months.

PERIOD_DIFF(NUMERIC, NUMERIC) -> LONG

QUARTER

Returns the quarter (1-4) for a given date/datetime.

QUARTER(TIME) -> LONG

SEC_TO_TIME

Takes a number of seconds and returns them formatted as a time (HH:MM:SS).

SEC_TO_TIME(NUMERIC) -> STRING

SECOND

Returns the seconds part of a date time value.

SECOND(TIME) -> LONG

STR_TO_DATE

Parses a string into a date. Returns a STRING if the format equals '%T' or starts with '%H'. Otherwise returns a DATETIME if the format string is longer than 10 characters, and returns a DATE in all other cases.

STR_TO_DATE(STRING input, STRING format) -> TIME|STRING

SUBDATE

Subtracts an interval from a date/datetime. Returns a date if the input is a date and the interval is in days or greater, otherwise returns a datetime.

SUBDATE(TIME, INTERVAL) -> TIME

SUBTIME

Subtracts a number of seconds from a date/datetime. The seconds can be decimal numbers to represent sub-second values.

SUBTIME(TIME, NUMERIC) -> DATETIME

SYSDATE

Returns the current date time.

SYSDATE() -> DATETIME

TIME

Returns the time portion of a datetime as a string formatted HH:MM:SS.

TIME(TIME) -> STRING

TIME_FORMAT

Formats the time portion of a datetime as a string.

TIME_FORMAT(TIME, STRING format) -> STRING

TIME_TO_SEC

Convert a time portion of a datetime to seconds.

TIME_TO_SEC(TIME) -> LONG

TIMEDIFF

Return the time difference between two time/datetimes as a HH:MM:SS string.

TIMEDIFF(TIME, TIME) -> STRING

TIMESTAMP

Returns a datetime value based on the arguments. Input should either be a date or a valid datetime string.

TIMESTAMP(TIME) -> DATETIME

TO_DATE

Converts an input into a DATE value.

TO_DATE(TIME) -> DATE

TO_DAYS

Return the number of days between a date expression and the year 0.

TO_DAYS(TIME) -> LONG

UNIX_TIMESTAMP

Returns the seconds since epoch (unix timestamp for a given date).

UNIX_TIMESTAMP(TIME) -> LONG

WEEK

Returns the week of year for a given date. Optionally a second "mode" parameter can be passed to indicate what starting day of the week to use and whether or not to use iso weeks.

The option parameter is a two digit number.

* First digit 1 or 2. 1 for January first based weeks, and 2 for the ISO standard weeks

* Second digit for the first day in the week Sunday - 1, Monday - 2, etc

Currently only 11 and 22 are supported.

WEEK(TIME[, LONG mode]) -> LONG

WEEKDAY

Return the day number in week for a given date/datetime. The return value is the same as DAYOFWEEK.

WEEKDAY(TIME) -> LONG

YEAR

Return the current year for a date/datetime. An optional mode parameter may also be passed, see WEEK.

YEAR(TIME[, LONG mode]) -> LONG

YEARWEEK

Return the year and week number as YYYYWW. An optional mode parameter may also be passed, see WEEK.

YEARWEEK(TIME[, LONG mode]) -> LONG

DATE_FORMAT

Formats a date into a string value.

DATE_FORMAT(TIME, STRING format) -> STRING

INITCAP

Capitalizes the first letter of each word. Words are delimited by white space or characters that are not alphanumeric.

INITCAP(STRING) -> STRING

INSTR

Returns the position (1 based) of a search string in another string, or zero if the string is not found.

INSTR(STRING source, STRING search_string) -> STRING

LEFT

Returns X characters from the start of a string.

LEFT(STRING, LONG) -> STRING

LENGTH

Returns the length of a string.

LENGTH(STRING) -> LONG

LOWER

Returns a string with all letters lowercased.

LOWER(STRING) -> STRING

LPAD

Left-pads a string with another string, to a certain length.

LPAD(STRING, LONG total_length, STRING pad_string) -> STRING

OCTET_LENGTH

Returns the number of bytes in the string.

OCTET_LENGTH(STRING) -> STRING

REPLACE

Replaces all occurrences of a substring in a string with a new string.

REPLACE(STRING, STRING from, STRING to) -> STRING

RIGHT

Returns X characters from the end of a string.

RIGHT(STRING, LONG) -> STRING

RPAD

Right-pads a string with another string up to X total characters.

RPAD(STRING, LONG total, STRING pad_string) -> STRING

SPLIT_PART

Splits a string by a given character and returns the requested part (1 based).

SPLIT_PART(STRING, STRING delimiter, LONG part)

SUBSTRING

Returns a substring of a specified length, starting at start_pos (1 based).

SUBSTRING(STRING, LONG start_pos, LONG length) -> STRING

TO_CHAR

Converts an expression to a string based on the passed in format.

TO_CHAR(ANY, STRING format) -> STRING

TO_NUMBER

Converts a string into a number. An optional pattern can be passed in. If the pattern doesn't include decimals the result will be LONG, otherwise the result will be a DOUBLE.

TO_NUMBER(STRING, STRING pattern) -> NUMERIC

TRIM

Trims whitespace from the begining and end of a string.

TRIM(STRING) -> STRING

UPPER

Converts all characters in a string to uppercase.

UPPER(STRING) -> STRING

APPROXIMATE_MEDIAN

Calculates an approximate median for a group of numerical values. The error tolerance may vary between engines.

APPROXIMATE_MEDIAN(NUMERIC) -> DOUBLE

APPROXIMATE_PERCENTILE

Calculates an approximate percentile for a group of numerical values. The percentile value must between 0 and 1. The error tolerance may vary between engines.

APPROXIMATE_MEDIAN(NUMERIC, DOUBLE percentile) -> DOUBLE

AVG

Calculates the average of a group of numbers.

AVG(NUMERIC) -> DOUBLE

COUNT

Counts the number of rows in the result set. If the DISTINCT keyword is used it counts the number of rows with distinct values for the passed in expression.

COUNT([DISTINCT] ANY) -> LONG

LISTAGG

Transforms non-null values from a group of rows into a list of values that are delimited by a configurable separator. A max_length can be supplied with an optional default value if the max_length is exceeded.

LISTAGG(ANY[, LONG max_length[, STRING separator[, STRING on_overflow]]]) -> STRING

MAX

Returns the maximum value in the result set group for the expression.

MAX(ANY) -> ANY

MEDIAN

Returns the median value in the result set group for the expression. Only works on numeric types.

MEDIAN(NUMERIC) -> NUMERIC

MIN

Returns the minimum value in the result set group for the expression.

MIN(ANY) -> ANY

STDDEV

Returns the statistical sample standard deviation in the result set group for the expression. Only works on numeric types.

STDDEV(NUMERIC) -> DOUBLE

STDDEV_POP

Returns the statistical population standard deviation in the result set group for the expression. Only works on numeric types.

STDDEV(NUMERIC) -> DOUBLE

SUM

Returns the sum of all values in the result set group for the expression. Only works on numeric types.

SUM(NUMERIC) -> NUMERIC

VAR_POP

Returns the population variance in the result set group for the expression. Only works on numeric types.

VAR_POP(NUMERIC) -> DOUBLE

CORR

Returns the coefficient of correlation in the result set group for the expression. Only works on numeric types.

CORR(NUMERIC) -> DOUBLE

FIRST_VALUE

Returns the first value in the partition for a given expression. If a window order by clause is not specified the value returned is not guaranteed to be consistent.

FIRST_VALUE(ANY) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> ANY

LAG

Returns the value of an expression in previous rows based on the offset provided (default 1). Optionally the user can supply a default value if the offset falls out of the range of the partition.

LAG(ANY[, LONG[, ANY]]) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> ANY

LEAD

Returns the value of an expression in subsequent rows based on the offset provided (default 1). Optionally the user can supply a default value if the offset falls out of the range of the partition.

LEAD(ANY[, LONG[, ANY]]) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> ANY

LAST_VALUE

Returns the last value in the partition for a given expression. If a window order by clause is not specified the value returned is not guaranteed to be consistent.

LAST_VALUE(ANY) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> ANY

MAX

Returns the maximum value of an expression within a partition. An order by clause can be supplied and if so the max will be calculated only with values previous to the current row based on the ordering.

MAX(ANY) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]])

MIN

Returns the minimum value of an expression within a partition. An order by clause can be supplied and if so the min will be calculated only with values previous to the current row based on the ordering.

MIN(ANY) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]])

NTILE

Equally divides an ordered data set (partition) into a {value} number of subsets within a window, where the subsets are numbered 1 through the value in parameter constant‑value. For example, if constant‑value= 4 and the partition contains 20 rows, NTILE divides the partition rows into four equal subsets of five rows. NTILE assigns each row to a subset by giving row a number from 1 to 4. The rows in the first subset are assigned 1, the next five are assigned 2, and so on.

If the number of partition rows is not evenly divisible by the number of subsets, the rows are distributed so no subset is more than one row larger than any other subset, and the lowest subsets have extra rows. For example, if constant‑value= 4 and the number of rows = 21, the first subset has six rows, the second subset has five rows, and so on.

If the number of subsets is greater than the number of rows, then a number of subsets equal to the number of rows is filled, and the remaining subsets are empty.

NTILE(LONG value) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> LONG

PERCENTILE

Simplified syntax for calculating a percentile. It takes in an expression and a percentile value and translates it into an analytic expression that is partitioned by the group by clause.

PERCENTILE(ANY expression, DOUBLE percentile) -> ANY

PERCENTILE_CONT

Returns the value that would fall into the specified percentile among a set of values in each partition.

NTILE(DOUBLE percentile) WITHIN GROUP ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> ANY

RANK

Ranks each row from 1 to partition row count based on the order clause in the over expression.

RANK() OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> LONG

STDDEV

Similar to Aggregate STDDEV but applies to a partition in the result set.

STDDEV(NUMERIC) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> DOUBLE

STDDEV_POP

Similar to Aggregate STDDEV_POP but applies to a partition in the result set.

STDDEV_POP(NUMERIC) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> DOUBLE

SUM

Sums the values in a partition. If an order by is included the sum will only include values at or preceding the current row according to that order.

SUM(NUMERIC) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> NUMERIC

VAR_POP

Similar to Aggregate VAR_POP but applies to a partition in the result set.

VAR_POP(NUMERIC) OVER ([PARTITION BY ANY] [ORDER BY ANY [ASC|DESC]]) -> DOUBLE