SQL & PL/SQL :: Converting Number Column To Date Column
Dec 25, 2012
I have a partitioned table with ~50 million rows that was setup with a number(10) instead of a date column. All the data in the table is ALWATS in this format YYYYMMDD
CREATE TABLE T1.monthly
(
SEQ_NUM NUMBER(10) NOT NULL,
DAY_DK NUMBER(10) NOT NULL
)
TABLESPACE USERS
PCTUSED 0
PCTFREE 10
[code]........
When I use the exchange partition method the parition is able to move the data from "monthly" table to "mth" table.
desc t1.mth; ### my temorary table
Name Null? Type
----------------------------------------- -------- ----------------------------
SEQ_NUM NUMBER(10)
DAY_DK NUMBER(10)
Than when I try to alter my temp table "mth". I get an error table must be empty to change column types.
alter table n546830.mth modify (DAY_DK date);
Next I tried making my temporary table "mth" a date column. When I an the exchange partition command I get the following error:
alter table t1.monthly exchange partition DEC_2012
with table t1.mth without validation;
alter table n546830.monthly exchange partition DEC_2012 with table n546830.mth without validation
*
ERROR at line 1:
ORA-14097: column type or size mismatch in ALTER TABLE EXCHANGE PARTITION
Method I can use to convert a number(10) to date column and keep the information in a table. Note, I don't care about HH:MM:SS as I never had that information to begin with and would be happy to set that part of the date column to all zeroes "00:00:00"
My requirment is to find out the period names and transactions which are in valid date formats and are less than sysdate and the non date formats are adjustments made by different users for their transactions
i have a varchar2 column containing string values that can be converted to date i.e. ('31-JUL-11') and that column also contains text strings in it. i.e. ('Some string data...')
records whose column value can be converted to date are extractable via where clause (i.e. those rows are associated with some fix number / flag)
now when i try to use to_date function i get the error that
" ORA-01858 a non-numeric character was found where a numeric was expected "
in sql i have added a where clause to only pick rows with flag, but even then it gives the error.
using a subquery in the from clause eliminates the error, but when i create it in a view it again gives the same error.
I have a column defined as Number( 8 ) which is supposed to have date values. I would like to check if all the rows in that table have valid dates. We could use to_date(coulmn_name, 'YYYYMMDD') and catch the rownums for error conditions using pl/sql. I would like to know if we could just do it using sql only and return the row numbers for those that are invalid dates?
I have one hirarchical query which return the parent to child hirarch level data. it has 11 level child data. i want to create column based on number of child in hirarchy. though i know it is 11 but it can change also.Is there any way i can create the column dynamically
We are using Oracle 11g with Apex 3.2 on AIX. We are reporting data from customer satisfaction surveys. I'm using the following sql to create my report
<code>select * from( select month,'Overall Satisfaction' as q_group, 1 as srt,Overall Satisfaction,site, case when count(*) < 31 then '*' else round((sum(ttos)/count(*))*100,0)||'/'|round((sum(bfos)/count(*))*100,0)||'/'||count(*) end ospct from v_XXX_report a,(select distinct month_dt month from v_XXX_report) b where Overall_Satisfaction is not null and year_dt = 2012
[Code]....
The problem is that site is not allways present and sometimes I have other variables in addition to site. This creates a situation where the month columns will not allways appear at the same column number. For example, When I run this query as is then the "JAN" column is Col3 (first column is a break, col2 is not shown). When I run this query without site then "JAN" is the second column. I would like to create column links for the "JAN" - "DEC" columns but not for any other columns.
Another issue - in the column link creation screen I can create up to 3 variables that I can pass to the next page. Since my query is a pivot I'm uncertain how to pass the column heading or the row value (for col2)
ie Overall Satisfaction JAN FEB MAR APR MAY ... Overall Satisfaction 12/12/200 12/12/210 12/12/220 12/12/230 12/12/240... Recommend 12/12/200 12/12/210 12/12/220 12/12/230 12/12/240... etc.
So if I clicked on the values at Recommend:FEB how can I get "Recommend" and "FEB" into variables that I can use on the next page? I've tried #column_name#, #month#, #q_name# and #APEX_APPLICATION.G_F10# but no luck.
I have a table where i have description column which free text column, the data in description column is seperated and i want to corvert 1 row data in multiple rows dependeing on the number of words.
eg
id description 78664 Pumps Alarm from CAMS RTU154
In the above example this column has 5 word so i want data in 5 rows like below
78664 Pumps 78664 Alarm 78664 from 78664 CAMS 78664 RTU154
This column data can be varied from 1 to any number of words.
is there some performance/access difference between a bitmap index on a number column and char(1) column? Both columns are not null with a default value.My application has a querie like this:
If I create a bitmap index on column "column_char", the access plan is not changed. But changing the column datatype to number(1) and obviously the values, the index is accessed and the cost decreases.This table has 4.000.000 rows. Oracle 11.2.0.2SO
There are millions of DISTINCT ID values in TABLE_1 and corresponding to each ID there are some values for all the days of a month. I need to insert these values in TABLE_2 in the above format.
need to create a table with single column by using select statement with multiple columns For Ex- i have 1 row with 10 columns (may be more than 10) like 'A','B','C','D','E','F','G','H',I','J' i written sql like select 'A','B','C','D','E','F','G','H','I','J' from dual
result is - 'A','B','C','D','E','F','G','H','I','J' with 10 columns
Now i need output lik this using SQL Text ------ 'A' 'B' 'C' 'D'
My requirement with one table which contains 20 columns. Now I want to convert 18 column into rows. I tried with the following query but its throwing sql error is given below.
select * from tmpl_pop_age_range_col_lvl UNPIVOT ( quantity FOR product IN (COUNT_0_TO_4 as 'a', COUNT_5_TO_9 as 'b', COUNT_10_TO_14 as 'c', COUNT_15_TO_19 as 'd', COUNT_20_TO_24 as 'e', COUNT_25_TO_29 as 'f',
[code].....
ERROR at line 4: ORA-00933: SQL command not properly ended
How can we partition a table based on date if it does not have a date column.
Actually I have to compare two tables on daily basis and fetch few rows from those two tables and enter it to a third table.But both these tables does not have a date column.
I am confused if i need to alter those tables and add date column or if there is some way in which i can compare the data from the two tables for that particular day only and not the whole table data.
I want to load data from a file using sqlldr.I have a table commissions ( technician_id char(5) , tech_name char(30) , Comm_rcd_date DATE , Comm_Paid_date DATE , comm_amt number(10,2) )
my file is 00001,TIMOTHY TROENDLY,2011-03-04T01:45:12+0006,2011-03-04T01:45:12+0007,123.56 00002,KENNETH KLEMENZ,2011-03-04T01:45:12+0006,2011-03-04T01:45:12+0009,123.56 00003,SHUNDAR ARDERY,2011-03-04T01:45:12+0006,2011-03-04T01:45:12+0005,123.56 write a ctl file to load this data.
my column type is NUMBER(10,0) ,it accept the input value from text field I using TO_NUMBER(?) to insert value into table, is the a way to handle if the input is 'aaaaaaaaaa' not digit?
In chapter 17 of the 'Oracle Database 10g, The Complete Reference' it says "You can increase or decrease the number of decimal places in a NUMBER column at any time". When trying to modify a column from NUMBER(38) to NUMBER(38,3) I get the ORA-01440 error. Is there a way around this when you have the maximum number of digits (precision = 38).
I am trying group by count on Monthly basis, entered_timestamp Column is of DATE datatype.
select count(*),to_char(entered_timestamp,'MM-YYYY') MONTHLY from payments where type = 'BOOK' group by to_char(entered_timestamp,'MM-YYYY') order by MONTHLY DESC
Above SQL doesn't gives the output in proper order ,
Can we have substr(date) as column name. I am trying to create a query where in I need to achieve something like this select s.xyz Today,s.abcd "6 Months ago" || to_char(add_months(sysdate,-6),'Mon rrrr') ||')'from sales_tab s.
I want update col1 whis is null to max(col1) ++ in a row, order by cr_date like 1,1,20110102 2,2,20110101 3,null,20110105 => 3,5,20110105 because this row is after 20110103 4,3,20110104 5,null,20110103 => 5,4,20110103 because this row is before 20110105
update test_table set col1 = (select max(col1) from test_table) + rownum where col1 is null;
