8.1 fixes great circle calculations!

8.1 fixes the great circle calculations in travel books. You can finally go from LAX to AKL without going over USA, Atlantic, Africa, Indian Ocean... Looking forward to finally ordering my travel books...
Radar-
rdar://6467664
Threads-
http://discussions.apple.com/thread.jspa?threadID=1886068

Now that is GOOD NEWS!
LN

Similar Messages

Can't show Greate Circle Route (in SRID 8307) in Map Viewer

I've got a task to show flight tracks on the world map in Map Viewer.
I've got a table with appropriate geometry
INSERT INTO USER_SDO_GEOM_METADATA (TABLE_NAME, COLUMN_NAME, DIMINFO, SRID) VALUES (
'TMP_LEG_ARC_001', 'LEG_GEOM',
SDO_DIM_ARRAY(
SDO_DIM_ELEMENT('LONG', -180.0, 180.0, 0.05),
SDO_DIM_ELEMENT('LAT', -90.0, 90.0, 0.05)
8307);
I've got a theme created from this column. It's evident that flights (and any other straight lines on the earth surface) should look like arcs but in MapViwer I can see straight lines.
It looks like Map Viewer just projected ends of line on the map and then draw straight line between them, instead of projecting whole line on the map (which should be converted into arc)
What have I done wrong?!
Edited by: Wishmaster on 08.11.2012 21:38

MapViewer doesn't display them as great circles. This will be fixed in a future release.
Meanwhile you can use Oracle Spatial functions to create the sissified lines which will approximate/be displayed as flight paths/great circles.
That is use something like Dan Geringer's (sample code so usual disclaimers apply) densify_line plsql stored procedure below.
The densify_line function takes two arguments:
• two_pt_line - The line to densify
• arc_tolerance - The interval (in meters) used to densify the line
The first argument calls for a two point line. It ensures the endpoints are included in the densified result.
You can also pass in a linestring with more than two points. The end points are ensured to be part of the densified line, but the intermediate shape points are not. To guarantee shape points, you can call densify_line iteratively with the 2 point line segments of a linestring.
CREATE OR REPLACE FUNCTION densify_line (two_pt_line MDSYS.SDO_GEOMETRY,
arc_tolerance number) RETURN MDSYS.SDO_GEOMETRY
DETERMINISTIC IS
lrs_segment MDSYS.SDO_GEOMETRY;
new_point MDSYS.SDO_GEOMETRY;
densified_line MDSYS.SDO_GEOMETRY := mdsys.sdo_geometry (two_pt_line.sdo_gtype,
two_pt_line.sdo_srid,
null,
mdsys.sdo_elem_info_array (1,2,1),
mdsys.sdo_ordinate_array ());
curr_measure NUMBER;
times_to_loop NUMBER;
BEGIN
IF two_pt_line IS NOT NULL
THEN
densified_line.sdo_ordinates.extend(2);
densified_line.sdo_ordinates(1) := two_pt_line.sdo_ordinates(1);
densified_line.sdo_ordinates(2) := two_pt_line.sdo_ordinates(2);
lrs_segment := SDO_LRS.CONVERT_TO_LRS_GEOM (two_pt_line);
times_to_loop := FLOOR (lrs_segment.sdo_ordinates(lrs_segment.sdo_ordinates.count) / arc_tolerance);
curr_measure := 0;
FOR i IN 1 .. times_to_loop LOOP
curr_measure := curr_measure + arc_tolerance;
densified_line.sdo_ordinates.extend(2);
new_point := sdo_lrs.locate_pt (lrs_segment, curr_measure);
densified_line.sdo_ordinates(densified_line.sdo_ordinates.count-1) := new_point.sdo_ordinates(1);
densified_line.sdo_ordinates(densified_line.sdo_ordinates.count) := new_point.sdo_ordinates(2);
END LOOP;
IF curr_measure < lrs_segment.sdo_ordinates(lrs_segment.sdo_ordinates.count)
THEN
densified_line.sdo_ordinates.extend(2);
densified_line.sdo_ordinates(densified_line.sdo_ordinates.count - 1) :=
two_pt_line.sdo_ordinates(two_pt_line.sdo_ordinates.count - 1);
densified_line.sdo_ordinates(densified_line.sdo_ordinates.count) :=
two_pt_line.sdo_ordinates(two_pt_line.sdo_ordinates.count);
END IF;
END IF;
RETURN densified_line;
END;
/

Least square circle calculation on a series of data points

I am taking a series of (x,y) part measurements that describe a circle (radius & rotation angle). The center of the part and center of rotation are offset. Looking for code to do a least square circle calculation on this data set so diameter and out-of-roundness can be estimated. Platform LabView 8.2.

Here is my simple solution from the above mentione thread:
(Is it not exaclty "least squares" in the classical sense but the sphere fit works typically very well..)
LabVIEW Champion . Do more with less code and in less time .

Fixed 30 Days calculation for any month..Not working fine when Split Occurs

Hello Everybody,
I am stuck with an issue..which is creating problem when there is a split..in Normal Conditions it's working fine...Kindly help me out on the same.
We have to calculate the Salary on 30 Days fixed basis and not on calender days...
eg For January Salary will be calculated based on 30 days and not on 31 days and for Feb also on 30 days not on 28/29 days..... Every thing is working fine for normal payroll i.e without any splits..
 But if there are any splits in the month then calculation for 2nd split goes wrong ....
 Eg. person is hired on 02feb 2007 and he got re-org on 17 feb 2007 with basic as rs.10000
As per Clients requrement it should be calculated as :
for period 2nd Feb 2007 - 16th Feb 2007 : 10000 * 15/30 : Working Fine in SAP
for period 17th Feb 2007 - 28th Feb 2007 : 10000* 14/30 : Not Working Properly In SAP calculation is done as 10000*12/30
We want it to calculate as 10000* 14/30
Kindly let me know wat can be done :
1 How can i write PCR..if that needs to be change
2. If user exit then how shall i go about the same
if some body has any 3rd option let me know...
Regards,
Ketal Parikh

Hi ,
Solved the issue by writing rule.
Regards,
Ketal
Edited by: Sikindar on Jul 3, 2010 10:28 PM

Fixed base overhead calculation on output quantity

Dear Guru's
I want to calculate the overhead @ fixed rate on the basis of output quantity of finished goods.
For the said purpose, I am using quantity based overhead calculation in SPRO. The overhead rate "C110" is being used with dependency D010 (Overhead Type/Overhead Key). Here i have maintained the fixed amount per tonne basis with overhead key "SAP1".
I have maintained the overhead key SAP1 in material master's. But still it's not calculating and not displaying anything in the line items. Also to my surprise, no error is displayed at the time of production order and standard cost estimate.
Please help me on the above issue.
A detailed help stating what possible errors I may have done would be helpful.
Thanks,
Sumit

Hi
Lets assume you choose Output (KG) to absorb the overhead
1. You need to create an Activity Type called PRDOVH and assign it in the PP Work center... Assign the cost element of this act type to a separate cost component in OKTZ
2. While creating the Routing, the Activity Qty = Operation Qty for this Act Type... i.e. If you are creating routing for 100 Units, specify the Act Qty for PRDOVH as 100...
3. The Formula in the work center will be fully variable for this Act Type.. i.e. the Act Qty should change in proportion to the Qty Confirmed....
4. The PP Guys will confirm the operation in PP... This Act Type will also get confirmed along with it
5. Now calculate Act Price KSII as usual and run revaluation... From this point onwards, follow the normla procedure which you follow in month end
br, Ajay M

Calculating distance between Lat Long points

I'm wondering if anyone out there has done this before in LabVIEW. Does anyone have a model of the "Great Circle" calculation? I have a file of lat/long points that I need to calculate the distance between. Any help??
V/r,
Chris

I haven't done it in LV, but if you go to Wikipedia and search for Great Circle Distance you'll find the formual and an example.

Distance between zipcodes

Want to find distance between two zipcodes that have their latitude and longitude stored in a table.
The table is as follows
CREATE TABLE distance (zipcode VARCHAR2, LNG NUMBER, LAT NUMBER)
I couldn't come up with a calculation or understand the mathematical calculation on line.. Can you help me with some stored procedure or straight query that will do..?
Thanks

Hi
Your best bet here is to convert the long/lat values to the spatial 'sdo_geometry' and use the sdo_geom.sdo_distance function.
-- create table
CREATE TABLE distance (zipcode VARCHAR2(5), LNG NUMBER, LAT NUMBER);
-- Add sample rows
insert into distance values (90210,-118.4099,34.0925);
insert into distance values (90806,-118.1894,33.8039);
-- add geometry column
alter table distance add geom sdo_geometry;
-- convert coordinates to geometry
update distance set geom = sdo_geometry(2001,8265,sdo_point_type(lat,lng,null),null,null);
commit;
-- example usage
select from_zipcode,to_zipcode,sdo_geom.sdo_distance(from_geom,to_geom,0.005,'unit=mile') distance
from
(select zipcode from_zipcode,geom from_geom
from distance
where zipcode = '90210') f,
(select zipcode to_zipcode,geom to_geom
from distance
where zipcode = '90806') t;
FROM_ZIPCODE TO_ZIPCODE DISTANCE
90210 90806 17.9768483148904 This distance is of course 'straight line' and not driving / walking time!
The '8265' above is the spatial reference identifier and I think is the right value to use here, although I'm sure my American counterparts will be happy to correct me if I have this wrong (i'm based in the UK and don't use American data that much anymore....)
If you wish to do any other analysis (such as nearest neighgbour etc) you will need to add metadata and a spatial index to this example....
You could implement the Great Circle calculation yourself and completely avoid using locator/spatial altogether but this would involve more code (= time = money).
Steve

Dynamic Calc processor cannot lock more than [100] ESM blocks during the calculation, please increase CalcLockBlock setting and then retry(a small data cache setting could also cause this problem, please check the data cache size setting).

Hi,
Our Environment is Essbase 11.1.2.2 and working on Essbase EAS and Shared Services components.One of our user tried to run the Cal Script of one Application and faced this error.
Dynamic Calc processor cannot lock more than [100] ESM blocks during the calculation, please increase CalcLockBlock setting and then retry(a small data cache setting could also cause this problem, please check the data cache size setting).
I have done some Google and found that we need to add something in Essbase.cfg file like below.
1012704 Dynamic Calc processor cannot lock more than number ESM blocks during the calculation, please increase CalcLockBlock setting and then retry (a small data cache setting could also cause this problem, please check the data cache size setting).
Possible Problems
Analytic Services could not lock enough blocks to perform the calculation.
Possible Solutions
Increase the number of blocks that Analytic Services can allocate for a calculation:
Set the maximum number of blocks that Analytic Services can allocate to at least 500.
If you do not have an $ARBORPATH/bin/essbase.cfg file on the server computer, create one using a text editor.
In the essbase.cfg file on the server computer, set CALCLOCKBLOCKHIGH to 500.
Stop and restart Analytic Server.
Add the SET LOCKBLOCK HIGH command to the beginning of the calculation script.
Set the data cache large enough to hold all the blocks specified in the CALCLOCKBLOCKHIGH setting.
Determine the block size.
Set the data catche size.
Actually in our Server Config file(essbase.cfg) we dont have below data added.
CalcLockBlockHigh 2000
CalcLockBlockDefault 200
CalcLockBlocklow 50
So my doubt is if we edit the Essbase.cfg file and add the above settings and restart the services will it work? and if so why should we change the Server config file if the problem is with one application Cal Script. Please guide me how to proceed.
Regards,
Naveen

Your calculation needs to hold more blocks in memory than your current set up allows.
From the docs (quoting so I don't have to write it, not to be a smarta***:
CALCLOCKBLOCK specifies the number of blocks that can be fixed at each level of the SET LOCKBLOCK HIGH | DEFAULT | LOW calculation script command.
When a block is calculated, Essbase fixes (gets addressability to) the block along with the blocks containing its children. Essbase calculates the block and then releases it along with the blocks containing its children. By default, Essbase allows up to 100 blocks to be fixed concurrently when calculating a block. This is sufficient for most database calculations. However, you may want to set a number higher than 100 if you are consolidating very large numbers of children in a formula calculation. This ensures that Essbase can fix all the required blocks when calculating a data block and that performance will not be impaired.
Example
If the essbase.cfg file contains the following settings:
CALCLOCKBLOCKHIGH 500 CALCLOCKBLOCKDEFAULT 200 CALCLOCKBLOCKLOW 50
then you can use the following SET LOCKBLOCK setting commands in a calculation script:
SET LOCKBLOCK HIGH;
means that Essbase can fix up to 500 data blocks when calculating one block.
Support doc is saying to change your config file so those settings can be made available for any calc script to use.
On a side note, if this was working previously and now isn't then it is worth investigating if this is simply due to standard growth or a recent change that has made an unexpected significant impact.

Using two different calculation type for one condition in purchasing

hi
purchasing price must be callculated as the mix of fixed and quantity calculation types, example is below;
for material A whose purchasing unit of measure is quantity (piece) but base unit of measure is gr.
1.from 1 gr to 500 gr discount must be 50 usd as fixed,so 1 gr discount is the same with 499 gr discount.
2. from 501 gr up to x, 0,10 usd per gr
for example;
if quantity = 10 and total default gr from material 450, discount must be 50 usd as fixed.
but during good receipt in migo 10 piece can be 600 gr so, price must be calculated as 600*0,10 = 60 usd.
or vice versa,in purchasing order total gram can be gretaer than the gr which will be used in migo during good receipt .
if quantity = 10 and total default gr from material 600, discount must be 60 usd as scaled.
but during good receipt in migo 10 piece can be 450 gr so, price must be calculated as 50usd as fixed.
it s a little complicated.
Thanks in advance.

TW Typewriter
If the coding is not written correctly, it would definitely affect the system performance. This is not only for user exit but also will be applicable to routines. So it makes no difference.
And is not difficult to get an access key to do this development
Not in all cases. In some projects, getting an Access Key itself is very difficult and it undergoes cumbersome process from client side. So we have to take a decision on case to case basis whether to go with routine or user exit.
G. Lakshmipathi

Commitment calculation partial reception of Services PO

Due to the Public Sector Management implementation, we changed our customizing setting it like this in SPRO and CUNI transaction:
CUNI:
Application parameters:
x Commercial meas. unit
SPRO:
Material Management - External Services Management - Source Determination and Default values - For Client:
x Service line level commitments.
This solved our issues for documents generated after the change of configuration, before this, commitments were incorrectly consumed entirely when posting a partial reception of services purchase orders.
The problem we are having right now is that we are identifying different behaviors in the commitment calculation between:
1) Purchase orders generated before changing customizing.
2) Purchase orders generated after the change in customizing.
In the first one, the partial reception reduced entirely the purchase order commitment (table COOI) and in the second one commitment was calculated correctly after a partial reception.
Additionally, we tried to fix the commitment calculation by using the RKANBU01 program, but it does not work as long as we maintain the unit of measurement u2018EAu2019 as u2018Commercial measurement unitu2019 in CUNI transaction.
Can somebody help me understand why this happens, meaning why does the system behave differently in these two cases having the same setting in customizing right now.
Thanks in advance.
Kind Regards!

Hi,
If all 5 services line item are include in single service entry sheet document, then you can't delete single service line item.
But your service entry sheet documents are different for all service line item then you can delete any one service entry sheet with ML81N.
Note: - Invoice documents is not generated against your service entry sheet documents.
Regards,
Mahesh Wagh

BPC75NW - Periodic Amount Not Calculating for records with blank YTD value

We are on BPC75NW SP6 and have a YTD consolidation application. Data is loaded on a monthly basis uising a process of importing data files from supplying systems and then loading them into BPC using the data manager append function with a transformation file. YTD amounts are as expected and 'most' periodic amounts are also calculated correctly via MDX mesures in reporting. However, in situations where we have a P&L YTD balance in one month (say January) and then in the next month (February) there is no longer a YTD balance in that P&L account, no record is loaded for February YTD and the system is not calculating any February periodic amount. Example:
January - Load $100 to a P&L Account which is the YTD amount
The January YTD amount = $100.
The January Periodic amount = $100.
February - Load no record to the same P&L account as above (as the YTD amount is now 0)
The February YTD amount in the system is blank
The February Periodic amount in the system is blank (This should be a NEGATIVE $100)
As a possible work around we have also tried to load a 0 to February. The data manager package shows the record as loaded successfully but since it is a 0 it really does not put it into the database.
Surely someone out there has already hit this situation and is willing to share the solution.
Thank you,
Vicki S.

This problem was resolved by installing SAP Note 1405993 - MDX: Error: Invalid MDX command with
CURRENTMEMBER. The current text on the note does not describe our error but SAP will not extend the text to cover this issue.
Prior to installing this note, the following other notes were also installed but they did not fix the periodic calculation issue. Under SAP's advisement we are going to leave them however as 'they should not have any side effects':
1529720 MDX: No data for calculated member and metadata reference
1590976 MDX: Incorrect data for IIF
1584196 MDX: Calculated members and different dimensions
1528374 MDX: System error GET_PRPTY_VALUE-02-
I will now close this message.
Vicki Shrontz

[svn] 3734: Fixing advanced CSS descendant selectors for the alpha release.

Revision: 3734
Author: [email protected]
Date: 2008-10-17 22:13:34 -0700 (Fri, 17 Oct 2008)
Log Message:
Fixing advanced CSS descendant selectors for the alpha release.
- We now check descendant types against any superclass.
- Removed StyleProtoChain.getClassStyleDeclarations()'s recursive "Class" lookup for each superclass of an IStyleClient during the search for type selectors in the hierarchy. This was replaced with a simple describeType-based description (called once per type).
- Added a method to report whether any advanced styles had been registered with the StyleManager to optimize style declaration matching for the default case of simple style declarations.
- Removed advanced selector cache key calculation from StyleProtoChain in favor of the legacy simple type level cache. Future work may be required to cache advanced, display list sensitive style information.
- Fixed the specificity calculation for global selectors.
QE: Yes
Doc: No.
Checkintests: Pass
mxunit styles tests: Pass
select mustella styles tests: Pass
Reviewer: Glenn
Bugs:
SDK-17351 - [Advanced CSS] Descendant selectors don't work for Application
SDK-17361 - [Advanced CSS] If you use descendant and id selectors together, the id selector doesn't work correctly
SDK-17385 - [Advanced CSS] If you use descendant and class selectors together, the class selector doesn't work correctly
Ticket Links:
http://bugs.adobe.com/jira/browse/SDK-17351
http://bugs.adobe.com/jira/browse/SDK-17361
http://bugs.adobe.com/jira/browse/SDK-17385
Modified Paths:
flex/sdk/trunk/frameworks/projects/framework/src/mx/core/UIComponent.as
flex/sdk/trunk/frameworks/projects/framework/src/mx/styles/CSSSelector.as
flex/sdk/trunk/frameworks/projects/framework/src/mx/styles/CSSStyleDeclaration.as
flex/sdk/trunk/frameworks/projects/framework/src/mx/styles/IAdvancedStyleClient.as
flex/sdk/trunk/frameworks/projects/framework/src/mx/styles/IStyleManager3.as
flex/sdk/trunk/frameworks/projects/framework/src/mx/styles/StyleManager.as
flex/sdk/trunk/frameworks/projects/framework/src/mx/styles/StyleManagerImpl.as
flex/sdk/trunk/frameworks/projects/framework/src/mx/styles/StyleProtoChain.as
flex/sdk/trunk/frameworks/projects/framework/src/mx/utils/NameUtil.as
flex/sdk/trunk/modules/compiler/src/java/flex2/compiler/css/StyleDef.vm
flex/sdk/trunk/modules/compiler/src/java/flex2/compiler/css/StyleModule.vm
flex/sdk/trunk/modules/compiler/src/java/flex2/compiler/mxml/gen/ClassDefLib.vm

HP 25 Calculator - Hard problem related to memory

Hi All,
I have an old HP 25 calculator, and I tried to revive it after 20 years have stopped.
There is a problem related to memory malfunction that is, the store functions and programe functions don´t work.
Can any one of you help me?
I would like to know if there is available some electrical scheme to understand and analyse what probrably is happening and also if there is some workshop here in São Paulo - Brazil, that could fix these old calculator.
I live in Brazil, in São Paulo city.
Thanks in advance.

Hi,
I think you will find the HP Calculator Museum forums a good place to ask:
http://www.hpmuseum.org/forum/
(ask your question in the General Forum).
I am sure there someone will give you some guidance.
Best regards.
Note: I do not work for HP, I just like playing with calculators :-)

Problem after upgradation from Oracle 8.1.7 to Oracle 9i

Hello ,
I currently installed Oracle 9i, and i m using Mapx 4.5,
and in the executable I am adding serverlayer
(a layer formed by querying the geoloc column from oracle
table), which was working fine with Oracle 8.1.7, but after
installation of Oracle 9i , i encounter the following error
while adding a serverlayer
ORA-13373 Element of type Extent is not supported for Geodetic
data
Cause: Element type extent for a polygon geometry is not
supported for
geodetic data.
Action: Convert the extent type polygon to a regular 5 point
polygon
and set
etype accordingly.
ORA-06512 at string line string
Cause: Backtrace message as the stack is unwound by unhandled
exceptions.
Action: Fix the problem causing the exception or write an
exception
handler
for this condition. Or you may need to contact your application
administrator or database administrator
Can anybody tell me what's going wrong ?
As I urgently got to fix it.
Thanks in advance
Binoy.

Hi,
Oracle9i has a geodetic data model. Your data has a geodetic
srid, which means that the data will need to conform to the rules
associated with geodetic data. These rules include:
No optimized rectangles are supported (my guess is the app is
using an optimized rectangle as a query window, but the data
could be the issue also).
No arcs or circle types are supported.
A single line segment cannot be equal to or greater than 1/2 the
great circle distance.
A polygon cannot have an area equal to or greater than 1/2 the
surface area of the earth.
You should try to find what version of Mapx supports Oracle9i.
Regards,
dan

How to make the objects got even distance in the Blend?

Wondering anyone has the same experience or not? I blend 2 circles together and adjust the spine from a straight line to wave curve line by adding points and adjusting the handles of the points. The space between the objects along the spine turn to uneven after that. I tried to use the "specified distance" in the option and add more points and modify the handles again. However, all in vain. Finally, I need to expand the blend and adjust the distance of those objects manually. Any suggestions to correct the stupid way? Thanks.

There is not really any way to get even spacing in a blend when the spine path has uneven curvature. (What the Specify Distance option does is take the length of the path, and divide it by the requested distance to determine the number of blend steps, but then once it has that number, they get spread out by exactly the same algorithm that is used when the number of steps is specified by a fixed number or calculated from the color difference.)
The length of the direction handles on a spine path affects the placement of objects, with longer direction handles pushing the blend steps away from the anchor point, and shorter direction handles pulling them in towards it. This tends to make the objects more tightly spaced when the path has sharp curves, and more loosely spaced where it is flatter.
If the direction handles are of equal length, and the distance between the two direction points is the same as their length, such as occurs on a circle, then the "pulls" balance out and the spacing ends up even. But when the curvature varies, so does the spacing.
This can be used to advantage to get non-linear gradations, by making a straight line and then pulling out direction handles that lie along the line. It can also be used to make objects on an elliptical orbit appear to be spaced in perspective, since they are closer together at the sharp ends of the ellipse and farther apart at the flat sides.
Often, however, what is desired is spacing that is independent of the curvature, such as you get with scatter brushes.
It was fully intended when live blends were implemented in AI 8 that even spacing would be an option, and would indeed be the default option, with logarithmic spacing and direction-handle controlled spacing being two other choices. The direction-handle controlled spacing was just implemented first not because it was most useful (it isn't) but because, believe it or not, it is
easiest. (Points on the interior of a bezier curve are natively addressed by t-value, which is a fraction of the direction handle distance. To obtain a position that is a fraction of the path length, one must first flatten the entire path into a polygonal approximation, then find the position by adding up the lengths of the flat segments. In contrast, one can obtain a point at a given fraction of the bezier t-value without flattening the path. This is the reason, by the way, that while scatter brushes have more even spacing between objects, they do not fit as tightly to the bezier as blend steps do - they are being placed on a straight line approximation of the path instead of directly on it.) The intention was to get the easy case of t-value based spacing working, and then add the harder case that requires flattening.
It turned out that we did not have time to add the extra options for AI 8, so it shipped as is. Then Macromedia sued us over live blends because Freehand had them first and as part of the settlement we had to promise to stop enhancing them. Now that the two companies have merged, that issue is moot, so enhancing blends is now a possibility for future versions of Illustrator. (All of you who complain about the cool features that Freehand has had for years and Illustrator hasn't, as if it were perverse and intransigent of Adobe not to follow suit, might want to keep in mind that copying the features of another company's software is often not allowed.)
In the meantime, scatter brushes can be a good alternative if all your blend steps are the same, but obviously they won't work if your blended shapes are changing along the path.

8.1 fixes great circle calculations!

Similar Messages

Maybe you are looking for