Common Functions
The system comes with function
COALESCE () : returns the first non-NULL value in the list, if all values in the list is NULL NULL is returned;
eg:
the concat () : string concatenation function;
eg:
Least () : returns the input parameters of a minimum
Greatest () : returns the largest of the input parameters (var1, var2 may bigint, double, datetime string or if all values are NULL NULL is returned.
Return value: the maximum value of the input parameters, the input parameters to return the same type when the implicit conversion is not present. NULL minimum. When the input parameter types are different, the comparison between the converted double double, bigint, string; string, comparing the converted datetime datetime. No other implicit conversion)
decode () : to achieve branch selection functionality
eg:select decode(customer_id,
1, 'Taobao',
2, 'Alipay',
3, Aliyun,
NULL, 'N/A',
'Others') as result
from sale_detail;
Decode the above following function implements the functions of if-then-else statement:
if customer_id = 1 then
result := 'Taobao';
elsif customer_id = 2 then
result := 'Alipay';
elsif customer_id = 3 then
result: = 'Aliyun;
...
else
result := 'Others';
end if;
if the function: if (logical condition, coumn1, coumn2) represents an output condition is satisfied, otherwise, the output value of the 2
eg:if(cap_direction not in('0','1'),null, cast(cap_direction as bigint));
substr () : Returns the string str substring from the beginning of length length start_position
eg: substr(""abc"", 2) = ""bc"";substr(""abc"", 2, 1) = ""b"";
TO_CHAR () : the Boolean type, bigint type, decimal type or double type represents the type of string into the corresponding
eg:to_char(123) = '123';to_char(true) = 'TRUE';to_char(1.23) = '1.23';to_char(null) = NULL;
TO_CHAR () : Datetime type, to convert the date value, if the input type string implicitly converted to datetime type involved in computing, other types of exception thrown.
eg:to_char(getdate(),'yyyymmdd')
the concat (coumn1, ',', coumn2) : string concatenation function
Match two precision:
concat(substr(to_char(lng),1,6),',',substr(to_char(lat),1,5)) like '120.08,30.28';
REGEXP_EXTRACT (coumn, '', Number): Split String Functions
Such as: Pro Road intersection with Vulcan toro
regexp_extract (inter_name, '(. *?) (Road)', 1) = Pro East
regexp_extract (inter_name, 'and (. *?) (intersection)', 1) = Vulcan toro
REGEXP_REPLACE : string replacement function
regexp_replace (round_name, '-', '', 1) represents a bar - replacing null
split_part string into functions
split_part ( 'North Central - Bridge dense', '-', 2) = density Bridge
InStr : calculating a substring positions in the string in str1 str2
instr('Tech on the net', 'e') = 2；instr('Tech on the net', 'e', 1, 1) = 2
cast
coors_convert (lng, LAT, 1): Google turn High German coors_convert (120.2334214,30.21829241,1)
WHERE judge_location(split_part(coors_convert(a.lng,a.lat,1),',',1),split_part(coors_convert(a.lng,a.lat,1),',',2))=1
Window function
统计 量: count, sum, avg, max / min, median, stddev, stddev_samp
排名：row_unmber,rank,dense_rank,percent_rank
Other categories: lag, lead, cluster_sample
--------------------
Basic usage; data into groups in accordance with a certain condition called window, each group is called a window
partition by column portion is used to specify the fenestration
The same partitioning column row value is considered in the same window a
How to specify the order by sorting the data in a window
Limitations: it can only appear in the select clause
Do not use the window function window functions and nested aggregate functions
And the same level of aggregation may not function together with
A odps sql statement can be used up to 5 window function
When Partition window, the same window contains up to 100 million rows of data
When the window with rows, x, y must be an integer greater than or equal to the constant 0, limit the scope 0-10000, value of 0 indicates the current row
You must use order by can specify the range of the window with rows ways
Not all window functions can be specified with the windows in rows, to support this use of window functions are avg, count, max, min, stddev and sum
----------------------
For chestnuts
select *,rank() over(partition by monitor_id order by distance) as mindistance_monitor_id from()
Custom Functions
Making the appropriate custom function based on Ali cloud odps
Note: this example because the odps version is too low: it is not used when creating aliyun example maven package step by step, instead of using their own package, due to the adoption of a step by step example to back out of the jar in question (out of the jar no class resources, profiles resources only).
Glossary:
UDF: user-defined scalar-valued function (user defined scalar function), which is one of the input and output relationship, reads one line of data (which may have a plurality of parameters), to write an output value
UDTF: Custom function table value (user defined table valued function), a function call is used to solve the multi-line output data for the scene, only to return a custom function a plurality of fields
UDAF: Custom aggregation function (user defined aggregation function), the input and output of many to one relationship, the plurality of input records aggregated into an output value (and can be coupled with group by statement)