View source for Data quality checks before upgrading to V6.0

{{Warning | All IMSMA databases are different and '''all''' of the steps below are important for getting all countries to a functional V6.0. }}
===Invalid / space coordinates (mandatory)===__NOEDITSECTION__
__FORCETOC__
{{TOC right}}

It was not possible to upgrade to V5.08.04 if there are invalid coordinates in the database. Please run this SQL [[Space coordinates|query]] anyhow because importing may overrun application rules. If you get any rows here, there are more SQL queries that you may run which will help you identify which object has the invalid coordinates.

===MGRS coordinates (mandatory)===__NOEDITSECTION__
In V5.xxx it was possible to enter MGRS with wrong format. The first [[MGRS coordinates|query]] checks for coordinates written with lower-case letters e.g. 4qfj12345678 and the second query checks for blanks e.g. 4 QFJ 1234 5678.

===Coordinates display rules (mandatory)===__NOEDITSECTION__
In IMSMA NG there are 3 rules on how to handle and display coordinates, based on how they were entered; 
* MGRS, 
* Distance/Bearing 
* X and Y. 

When data has been migrated or imported it is possible that the column userinputformat in the table ''geopoint'', which controls coordinate display, has not been set correctly. When the display field is set correctly the columns are high-lighted with yellow. If the user tries to edit a row which does not have one of the rules set, the coordinates will not be populated in the Point window. This issue only affects the edit of Auxiliary data coordinates and Data Entry Forms in the Workbench [[Coordinate display rules|(query)]].

The three fields coordsys, coordformat and userinputformat must be in a valid combination '''and''' must be consistent with the value of userenteredcoord e.g. (WGS 84, Degrees:Minutes:Seconds, X and Y) with 17.2345|25.5678|en|UK is not a valid combination. If these records are not corrected the area/length recalculation and the upgrade will not work [[Coordinate display rules|(query)]].

===Length of Point ID===__NOEDITSECTION__
Point (local) ID should be unique and clearly identify single and polygon points. But it is possible to type blanks instead of characters/numbers. If spaces have been used it is not possible for the users to see difference when entering/editing the points.
If you get length 0, 1 or 2 it is recommended to look into the values of Point ID [[Point ID Length|(query)]].

===Missing approval information (mandatory)===__NOEDITSECTION__
Due to changed behaviour of migration/import scripts in version 5.08.02 and a bug in 5.08.04 approval information may be missing. Note that ''Last updated'' gets updated e.g. when templates are switched and should not be used for setting Approved Date. If the Data Entry Form that misses Approved Date is an Activity or Education then you might be able to use End date [[Missing approval information|(query)]].

{| class="wikitable"
|-
! Column !! Old label !! New label
|-
| reportReceivedDate|| Initiated || Data Entry Date
|-
| reportCompletedDate|| Submitted || Submitted Date
|-
| reportVerifiedDate|| Approved || Approved Date
|-
| dateofreport|| Date of Report || Date of Information
|-
| dataentrydate|| Last updated || (no change)
|-
|}

===Missing Date of Information (mandatory)===__NOEDITSECTION__
Date of Information (stored in table ''fieldreport'') is used when {{IMSMANG}} calculates the Summary (former Current view). If this data is missing then the result may not be as the expected when the Summary is updated. Note that ''Last updated'' gets updated e.g. when templates are switched and should not be used for setting Date of Information [[Missing Date of Information|(query)]]. 

===Missing Local ID (mandatory)===__NOEDITSECTION__
In some database also Form ID and the items' local ID have been missing. The [[Missing Local ID|queries]] do not include Country structure, Organisation, Place, Task and WorkItem.

===Data Entry Form Templates===__NOEDITSECTION__
Some countries have many published templates that have never been used which makes it difficult to know which template to update etc. It is also good to know which templates that have been used for data entry in case you need to update or switch them [[SQL Templates|(queries)]].

===DIM categories (mandatory)===__NOEDITSECTION__
If categories have been deleted there will be errors when the upgrade script is applied. As a quick indication if there will be errors, numbers of categories may be used. There should be at least 52 categories.

These 5 categories for Task and Work Item should exists [[DIM Categories|(queries)]].

{| class="wikitable"
|-
! Parent !! Category
|-
| TASK || GENERAL_INFO
|-
| TASK || PLANNING_MONITORING
|-
|TASK	|| UNCATEGORISED
|-
| WORK_ITEM || GENERAL_INFO
|-
| WORK_ITEM ||	UNCATEGORISED
|-
|}

===Number of enumeration categories and values (mandatory)===__NOEDITSECTION__
If the country has deleted standard enums from the table ''imsmaenum'' then upgrade scripts will not give the expected result which will create a lot of problems. As a quick indication if there will be errors, numbers of categories and values may be used. There should be at least 117 categories and 978 values in the result set of these [[Enum categories and values|queries]].

===Duplicate enumvalues (mandatory)===__NOEDITSECTION__ 
Duplicate enumvalues cause import problems, see [[Duplicate Enumvalue]]. The upgrade scripts will create some duplicates so it is important to know if there were duplicates in 5.08.04 too ([[Duplicate Enumvalue SQL|query]]).

===Number of translations (mandatory)===__NOEDITSECTION__
If the country has deleted standard translations from the table ''translation'' then upgrade scripts will not give the expected result which will create a lot of problems. The result of this [[Translations|query]] should be at least 1269 (English) translations.

===CDF display format (mandatory)===__NOEDITSECTION__
The combination of cdf_datatype in table ''customdefinedfield'' and fieldtype in table ''field'' must be correct. When data type has been changed directly in table ''customdefinedfield'' or data has been migrated these combination might be wrong [[CDF display format|(query)]].

Common errors are. 
*display_mechanism for multi select  is set to RADIO_BUTTON;
* no display_mechanism is set for single select fields;
* wrong fieldtype for multi and single selects;
* display_mechanism is "" i.e. length is 0.

The combinations in the table below is the correct 5.08.04 values
{| class="wikitable"
|-
! fieldtype !! display_mechanism !! cdf_datatype
|-
| cdf || (null)|| DATE
|-
| cdf || (null)|| GAZETTEER
|-
| cdf || (null)|| MULTI_SELECT
|-
| cdf || (null)|| NUMBER
|-
| cdf || (null)|| ORGANISATION
|-
| cdf || (null)|| PLACE
|-
| select_cdf || COMBO_BOX || SINGLE_SELECT
|-
| select_cdf || RADIO_BUTTONS || SINGLE_SELECT
|-
| cdf || (null)|| TEXT_FIELD
|}

===CDF missing in table field (mandatory)===__NOEDITSECTION__
We have had a few cases where CDFs have been missing in the table ''field'' or in the table ''customdefinedfield'' [[CDF missing in field|(queries)]]. 

===Duplicate CDFs (mandatory)===__NOEDITSECTION__
In 5.08.04 it was allowed that CDFs had the same name and same parent if they had different data type. This [[Duplicate CDF|query]] checks if there are duplicates.

===Multi-select CDFs (mandatory)===__NOEDITSECTION__
There is a display issue with imported/migrated Multi-select CDFs in data entry templates. Manually entered data is not affected. Novetta is currently (2014-06-18) looking into a long-term solution for this. The short-term solution is a database update and to ask the country if they import/migrate regularly any multi-select CDFs[[Multi-select CDFs | (query)]].
 
===CDF never used for Data Entry===__NOEDITSECTION__
Note that a CDF may be included in Data Entry templates, Summary templates, Saved searches, etc. but have no values in table ''cdfvalue''. This is not an error but if they have many non-used CDFs it is an indication that IM procedures may be improved [[CDF not used|(query)]].

===Records with empty values in table cdfvalue===__NOEDITSECTION__
The next level of check is to check if the value ‘ ‘ have been stored AND if all different data types are NULL. An old bug created empty strings for specially Auxiliary data CDFs. This bug created unnecessary many rows in the table ''cdfvalue'' and makes it impossible to delete CDFs that actually never have been used for data entry [[Empty CDF|(query)]].

===Duplicated values in allow_value_set===__NOEDITSECTION__
Duplicates may have been by migration and import mistakes [[Duplicate allow value|(query)]]. This could be addressed when CDFvalues are moved to standard fields after upgrade.

===Invalid date in link table (mandatory)===__NOEDITSECTION__
The date value 0000-00-00 00:00:00 is invalid and has been found in the table ''link'' [[Link invalid date|(query)]].

===Attachment table (mandatory)===__NOEDITSECTION__
One country had changed the name of the column ''Filedescription''. Note that the name has to start with capital F.
Check this with Navicat and update the name if necessary in Table design [[Attachment SQL |(query).

===Orphans in Country Structure (mandatory)===__NOEDITSECTION__
If there are orphans in the country structure there will be problems e.g. with creating a staging area database [[Country structure orphans|(queries)]].

===Duplicates in the Country Structure (mandatory)===__NOEDITSECTION__
Having duplicates on name (on the same level/node) in the country structure e.g. two villages called “Berg” in the one municipality creates a lot of problems when importing data and creating statistics. Note that the duplicates are not shown in the GUI [[Country structure duplicates|(query)]].

===Accident - Unknown device (mandatory)===__NOEDITSECTION__
The field ''Unknown device'' is an old Yes/No field from Legacy. This field is very often conflicting or duplicate to the values in table ''accdeviceinfo'' and therefore it was decided to remove it from the database. The upgrade script will update the '''Type of Accident''' depending on the values of ''Unknown Device'', see the rules [[Unknown Device| here]]. Run the [[Unknown device Accident| query]] and highlight to the country that Accident type will be updated.

===Data quality queries - Duplicates===__NOEDITSECTION__
The upgrade script will add some of these [[SQL Saved search Duplicate ID|queries]] to the database as saved searches. You may run them now as SQL queries or later as Saved searches.

===Data quality queries - No geographical data===__NOEDITSECTION__
The upgrade script will add these [[SQL Saved search No geo data|queries]] to the database as saved searches. You may run them now as SQL queries or later as Saved searches.

===Data quality queries - No links===__NOEDITSECTION__
The upgrade script will add these [[SQL Saved search No links|queries]] to the database as saved searches. You may run them now as SQL queries or later as Saved searches.

===Data quality queries - Different Location===__NOEDITSECTION__
The upgrade script will add these [[SQL Saved search Different Location|queries]] to the database as saved searches. You may run them now as SQL queries or later as Saved searches.

===Data quality queries - Task===__NOEDITSECTION__
The upgrade script will add these [[SQL Saved search Task|queries]] to the database as saved searches. You may run them now as SQL queries or later as Saved searches.

===Data quality queries - Mine Action quality===__NOEDITSECTION__
The upgrade script will add these [[SQL Saved search MA quality|queries]] to the database as saved searches. You may run them now as SQL queries or later as Saved searches. There are more Saved Searches than these mentioned here added to the database but they are referencing to tables/fields that do not exist in 5.08.04.

[[Category:NoPublic]]

{{NavBox Upgrade}}