formatting a calculated number field on a form

My database consists of medical information and includes, specifically, the
age at which certain events may have occured. I have written VBA macro to
calculate the age.

For calculating, for example, the age at death, the arguments are the date of
birth (DoB) and the date of death (DoD). If a patient is alive, DoD is null,
and the VBA function returns a value of 0. If the patient has died, the age is
calculated in decimal years, rounded to 2 decimal places, e.g. something like
78.32.

I have created a query with a field that calculates ages as described above.
The expression used on the query for age at death is

AgeAtDeath: Age(DoB,DoD)

and it is formatted as Fixed.

The query is the data source for a form. I would like to achieve the following
appearance on the form:

1. For a live patient, I see 0.00. I want it to look blank.
2. For a dead patient that lived exactly 69 years, I see 69 rather than 69.00.
For a patient that lived 38.2 years, I see 38.2 rather than 38.20.

I can get around those 2 issues by changing the expression on the form to
something like

Iif(Age(DoB,DoD)=0,"",Format$(Age(DoB,DoD),"0.00") )

But now the field on the form contains text rather than a number, so there are
problems filtering correctly. If, for example, I want to see only patients who
died after age 40, I type in the Filter-by-form box

40

and Access changes that to

"40"

The consequence is that patients who died between the ages of 5.00 and 9.99
are incorrectly included since the filter is comparing text rather than
numbers.

Is there any way to achieve the desired format ('blank' or 2 decimal places)
while still having the field data treated as a number for the purpose of
filtering?

Correction: I said "changing the expression on the form...". I should have
said "changing the expression in the query..."


On Wed, 16 Nov 2005 23:22:24 -0600, Myrna Larson
wrote:

My database consists of medical information and includes, specifically, the
age at which certain events may have occured. I have written VBA macro to
calculate the age.

For calculating, for example, the age at death, the arguments are the date of
birth (DoB) and the date of death (DoD). If a patient is alive, DoD is null,
and the VBA function returns a value of 0. If the patient has died, the age
is
calculated in decimal years, rounded to 2 decimal places, e.g. something like
78.32.

I have created a query with a field that calculates ages as described above.
The expression used on the query for age at death is

AgeAtDeath: Age(DoB,DoD)

and it is formatted as Fixed.

The query is the data source for a form. I would like to achieve the
following
appearance on the form:

1. For a live patient, I see 0.00. I want it to look blank.
2. For a dead patient that lived exactly 69 years, I see 69 rather than
69.00.
For a patient that lived 38.2 years, I see 38.2 rather than 38.20.

I can get around those 2 issues by changing the expression on the form to
something like

Iif(Age(DoB,DoD)=0,"",Format$(Age(DoB,DoD),"0.00") )

But now the field on the form contains text rather than a number, so there
are
problems filtering correctly. If, for example, I want to see only patients
who
died after age 40, I type in the Filter-by-form box

40

and Access changes that to

"40"

The consequence is that patients who died between the ages of 5.00 and 9.99
are incorrectly included since the filter is comparing text rather than
numbers.

Is there any way to achieve the desired format ('blank' or 2 decimal places)
while still having the field data treated as a number for the purpose of
filtering?

Myrna Larson wrote:
I have created a query with a field that calculates ages as described above.
The expression used on the query for age at death is

AgeAtDeath: Age(DoB,DoD)

and it is formatted as Fixed.

The query is the data source for a form. I would like to achieve the following
appearance on the form:

1. For a live patient, I see 0.00. I want it to look blank.
2. For a dead patient that lived exactly 69 years, I see 69 rather than 69.00.
For a patient that lived 38.2 years, I see 38.2 rather than 38.20.

I can get around those 2 issues by changing the expression on the form to
something like

Iif(Age(DoB,DoD)=0,"",Format$(Age(DoB,DoD),"0.00") )

But now the field on the form contains text rather than a number, so there are
problems filtering correctly. If, for example, I want to see only patients who
died after age 40, I type in the Filter-by-form box

40

and Access changes that to

"40"

The consequence is that patients who died between the ages of 5.00 and 9.99
are incorrectly included since the filter is comparing text rather than
numbers.

The principles of tiered architecture tells us to use the data layer to
return raw values and use the front end to apply any formatting. Access
tends to blur the lines between these boundaries e.g. a .mdb file is
referred to as and 'Access database' even though it contains front end
elements such as reports and forms.

Worse, I see evidence of users embedding hard coded references to front
end elements in the database layer e.g. a Query object (a.k.a. a VIEW),
which is very much a data layer element, containing a form's control in
the SQL where clause. Using a tiered approach, entirely possible in
Access, would be to create a Parameter Query object (a.k.a. a
PROCEDURE, stored proc, etc) in the data layer with strongly typed
parameters (possibly supplying default values); the front end would
then invoke the Query/proc passing values as required.

Why is this tiered approach not favored in Access? I really have no
idea. I get the impression that most Access font end developers choose
to have forms with bound controls: always bound, always connected, the
data is immediately committed as soon as a control's value is changed,
etc. What criteria do they use to make this choice: default behavior,
less/no programming required, lack of/no specific training in computing
science, quick kludge due to time/cost constraints? Again, I have no
idea.

To answer your question less obliquely, if you can't/won't change to a
tiered approach, how about having the Query/VIEW return two columns:
one being the raw data which is kept hidden from the user but available
to the front end developer under the hood in the form, and one column
formatted for display purposes only.

Are you suggesting that I should not use the built-in Filter-by-Form to allow
the user to set the filter criteria, but that I should develop my own form for
that purpose?

On 17 Nov 2005 02:15:37 -0800, wrote:


Myrna Larson wrote:
I have created a query with a field that calculates ages as described
above.
The expression used on the query for age at death is

AgeAtDeath: Age(DoB,DoD)

and it is formatted as Fixed.

The query is the data source for a form. I would like to achieve the
following
appearance on the form:

1. For a live patient, I see 0.00. I want it to look blank.
2. For a dead patient that lived exactly 69 years, I see 69 rather than
69.00.
For a patient that lived 38.2 years, I see 38.2 rather than 38.20.

I can get around those 2 issues by changing the expression on the form to
something like

Iif(Age(DoB,DoD)=0,"",Format$(Age(DoB,DoD),"0.00") )

But now the field on the form contains text rather than a number, so there
are
problems filtering correctly. If, for example, I want to see only patients
who
died after age 40, I type in the Filter-by-form box

40

and Access changes that to

"40"

The consequence is that patients who died between the ages of 5.00 and 9.99
are incorrectly included since the filter is comparing text rather than
numbers.

The principles of tiered architecture tells us to use the data layer to
return raw values and use the front end to apply any formatting. Access
tends to blur the lines between these boundaries e.g. a .mdb file is
referred to as and 'Access database' even though it contains front end
elements such as reports and forms.

Worse, I see evidence of users embedding hard coded references to front
end elements in the database layer e.g. a Query object (a.k.a. a VIEW),
which is very much a data layer element, containing a form's control in
the SQL where clause. Using a tiered approach, entirely possible in
Access, would be to create a Parameter Query object (a.k.a. a
PROCEDURE, stored proc, etc) in the data layer with strongly typed
parameters (possibly supplying default values); the front end would
then invoke the Query/proc passing values as required.

Why is this tiered approach not favored in Access? I really have no
idea. I get the impression that most Access font end developers choose
to have forms with bound controls: always bound, always connected, the
data is immediately committed as soon as a control's value is changed,
etc. What criteria do they use to make this choice: default behavior,
less/no programming required, lack of/no specific training in computing
science, quick kludge due to time/cost constraints? Again, I have no
idea.

To answer your question less obliquely, if you can't/won't change to a
tiered approach, how about having the Query/VIEW return two columns:
one being the raw data which is kept hidden from the user but available
to the front end developer under the hood in the form, and one column
formatted for display purposes only.

Myrna Larson wrote:
Are you suggesting that I should not use the built-in Filter-by-Form to allow
the user to set the filter criteria, but that I should develop my own form for
that purpose?

One step at a time: you need the data layer elements before you can
build the front end elements (didn't you say the built in features
weren't working for you?)

You've given no details about your base tables, constraints, business
rules for age calculations, etc so I'll fake it with this test table:

CREATE TABLE Test (
key_col INTEGER NOT NULL UNIQUE,
DoB DATETIME NOT NULL,
DoD DATETIME
)
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (1, CDATE('1970-01-01'), CDATE('2005-11-01'))
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (2, CDATE('1940-02-02'), CDATE('2005-02-01'))
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (3, CDATE('1980-03-03'), NULL)
;

I'll fake the calculation to suit, too. My formatting should be as per
your spec, though. If you are not using 'ANSI SQL-92 query mode' or
ADO, you may find it quicker to paste the following SELECT into a new
Query object and save it manually:

CREATE VIEW TestView
AS
SELECT key_col,
IIF(DoD IS NULL, 0, DATEDIFF("d", DoB, DoD) / 365.25) AS age_at_death,
IIF(age_at_death = 0,
'',
FORMAT$(age_at_death,
IIF(RIGHT$(FORMAT$(age_at_death, '##0.00'), 2) = '00',
'##0',
'##0.00'
)
)
) AS display_age_at_death
FROM Test
;

Now we need another element in the data layer which will take a
strongly typed parameter for age (I assume you want to query on whole
years) and the applicable operators can be encoded. Again, if you want
to do this manually you can use a SELECT in a Query object using the
PARAMETERS declaration but you will have to forego the default
parameter value:

CREATE PROCEDURE TestProc (
:value INTEGER,
perator_code INTEGER = 1
) AS
SELECT key_col, age_at_death, display_age_at_death
FROM TestView
WHERE SWITCH(
age_at_death = 0, 0,
perator_code = 1, IIF(age_at_death = :value, 1, 0),
perator_code = 2, IIF(age_at_death :value, 1, 0),
perator_code = 3, IIF(age_at_death :value, 1, 0),
perator_code = 4, IIF(age_at_death = :value, 1, 0),
perator_code = 5, IIF(age_at_death = :value, 1, 0),
perator_code = 6, IIF(age_at_death :value, 1, 0),
TRUE, 0) = 1
;

Now, to query on 40, use

EXEC TestProc 40, 3

With the required data layer elements in place, you can collect the
parameter values using forms and use them to call the proc/Parameter
Query.

This is an MDB database, not SQL. There is only 1 table (it's really just a
small flat-file which shouldn't require a relational DB, but see comment
below). There are just 700 patients.

Some persons have a genetically determined disease, others do not. Of the
affected persons, some have been treated with surgery, others have not. So
basically there are 3 groups: unaffected, affected but not treated, and
affected and treated.

We are tracking longevity and the occurrence and timing of heart attacks and
strokes in the 3 groups. I have created all of the queries using the QBE grid;
I haven't needed parameter queries. The form in question is used primarily for
data entry but on occasion I want to filter by age as well.

Each record contains data about a single patient, about 32 fields. There are 5
date/time fields: date of birth, date of surgery, date of heart attack or
stroke, date of last contact, and date of death. Any or all of these fields
can be null. I calculate 4 "ages": age at surgery, age at heart attack, age at
last contact, and age at death. For a living patient with no surgery and no
heart attack, only age at last contact is not null.

This data could easily be tracked as a flat-file in Excel, except for 3 Excel
problems: Excel can't handle dates prior to 1/1/1900 (some of these patients
were born in the 1800's), the user can fairly easily "scramble" the table by
improper sorting, and it's much more difficult to set up the "queries" and
filtering in Excel.

BUT... in Excel, I can easily create a data entry form and specify a number
format that (1) does not display zero values, (2) ALWAYS shows 2 decimal
places for non-zero values, and (3) achieving that format doesn't require
changing the type of the underlying data to text.

I don't seem to be able to do that in Access.

This is so EASY in Excel; I don't know why it should be so DIFFICULT
(impossible?) in Access. But I guess that's the way it is. Maybe the next
version of Excel will allow dates prior to 1/1/1900.

Anyway, thank you for your help.

PS: I have from a 25 year background in Basic programming and use of Excel. (I
was a VB and Excel MVP from 1994-2004.) But I am essentially a newbie when it
comes to Access.


On 18 Nov 2005 01:08:00 -0800, wrote:


Myrna Larson wrote:
Are you suggesting that I should not use the built-in Filter-by-Form to
allow
the user to set the filter criteria, but that I should develop my own form
for
that purpose?

One step at a time: you need the data layer elements before you can
build the front end elements (didn't you say the built in features
weren't working for you?)

You've given no details about your base tables, constraints, business
rules for age calculations, etc so I'll fake it with this test table:

CREATE TABLE Test (
key_col INTEGER NOT NULL UNIQUE,
DoB DATETIME NOT NULL,
DoD DATETIME
)
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (1, CDATE('1970-01-01'), CDATE('2005-11-01'))
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (2, CDATE('1940-02-02'), CDATE('2005-02-01'))
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (3, CDATE('1980-03-03'), NULL)
;

I'll fake the calculation to suit, too. My formatting should be as per
your spec, though. If you are not using 'ANSI SQL-92 query mode' or
ADO, you may find it quicker to paste the following SELECT into a new
Query object and save it manually:

CREATE VIEW TestView
AS
SELECT key_col,
IIF(DoD IS NULL, 0, DATEDIFF("d", DoB, DoD) / 365.25) AS age_at_death,
IIF(age_at_death = 0,
'',
FORMAT$(age_at_death,
IIF(RIGHT$(FORMAT$(age_at_death, '##0.00'), 2) = '00',
'##0',
'##0.00'
)
)
) AS display_age_at_death
FROM Test
;

Now we need another element in the data layer which will take a
strongly typed parameter for age (I assume you want to query on whole
years) and the applicable operators can be encoded. Again, if you want
to do this manually you can use a SELECT in a Query object using the
PARAMETERS declaration but you will have to forego the default
parameter value:

CREATE PROCEDURE TestProc (
:value INTEGER,
perator_code INTEGER = 1
) AS
SELECT key_col, age_at_death, display_age_at_death
FROM TestView
WHERE SWITCH(
age_at_death = 0, 0,
perator_code = 1, IIF(age_at_death = :value, 1, 0),
perator_code = 2, IIF(age_at_death :value, 1, 0),
perator_code = 3, IIF(age_at_death :value, 1, 0),
perator_code = 4, IIF(age_at_death = :value, 1, 0),
perator_code = 5, IIF(age_at_death = :value, 1, 0),
perator_code = 6, IIF(age_at_death :value, 1, 0),
TRUE, 0) = 1
;

Now, to query on 40, use

EXEC TestProc 40, 3

With the required data layer elements in place, you can collect the
parameter values using forms and use them to call the proc/Parameter
Query.

FWIW, I have found the answer to my question. It's what is called a "custom
number format" in Excel.

In Access a custom format has 4 sections, separated by semicolons, which
specify the formats for for positive numbers, negative numbers, 0, and Null,
respectively.

To get the appearance I want on the form, I use this format specification, but
without the quotes:

"0.00; ; ; "

This solves my problem completely!

The reason I couldn't get it to work in the past was because I typed the
format string without the spaces between the semicolons (meaning 'don't
display', which is how you do it in Excel), but Access changed that to
"Fixed".

Before you ask why I didn't use Help for this, I did. I searched for "custom
number format" and got nothing. The way to find it is, on the Property sheet
for the control, select Format, and press F1. There they describe the 4
sections and comment that "If you use multiple sections but don't specify a
format for each section, entries for which there is no format will either
display nothing or default to the formatting of the first section."

If they had done the former -- display nothing -- I wouldn't have had a
problem. But it's solved now. Thanks for your help.

Myrna Larson wrote:
This is an MDB database, not SQL. There is only 1 table

it's really just a
small flat-file which shouldn't require a relational DB

An .mdb is a flat file, that's why Access users talk terms of 'fields'
and 'records' rather than the relational terms 'columns' and 'rows'.
With Jet, however, you get a relational 'view' of the data in the file,
therefore an .mdb can be considered a DBMS (database management
system). Similarly, the .mdb can be considered a SQL because Jet
provides a SQL parser with which to query the data in the file; it fell
short of SQL-92 compliance but it's still pretty good and in some areas
even better than the SQL Server 2000 implementation (CASCADE paths and
CHECK constraints across multiple tables spring to mind). Speaking of
which, the SQL I posted was Jet SQL syntax and not that of SQL Server;
notice the parameter names: SQL Server forces you use a @ symbol
whereas Jet will let you use the Standard SQL colon character (so
that's another one in Jet's favour, then).

This is so EASY in Excel; I don't know why it should be so DIFFICULT
(impossible?) in Access.

I was a VB and Excel MVP from 1994-2004

Access is another (superfluous) layer on top of the .mdb and Jet. You
may find it helpful to use its GUI designers to build tables, VIEWs,
PROCEDURES. You certainly aren't compelled to use Access to build your
front end application; you could do this in Excel, in which you are
undoubtedly better skilled (I recognized the name, of course:
http://groups.google.com/group/micro...db13d97a74338c).
MS Access forms are quirky and unlike UserForms, Windows Forms and
every other forms engine I've encountered. They are a learning curve to
nowhere, IMHO.

This is really why I made the point about a tiered approach. Get the
database to do the things it does best i.e. manage data. I encourage
you to make your solution Access independent e.g. if you build your age
calculation in Access VBA and call it in an Access Query (VIEW) you
will not be able to use that Query/VIEW from Excel or any other front
end program; if the calculation can be build in Jet SQL then do so.

Maybe the next
version of Excel will allow dates prior to 1/1/1900.

I've been reading about the increase in some of the old limits
(http://blogs.msdn.com/excel/archive/...26/474258.aspx) but
haven't seen explicit mention of dates

I have been watching this thread, but since I would have headed in rather a
different direction from that taken by peregenem I have sat back and
observed the exchange. I'm not sure why peregnem posts here, since he
clearly dislikes Access so much.
Access is designed to use bound controls. If you are opposed to the concept
of bound controls, Access is probably not the way to go.
I'm not sure how (or if) you got the Age function to work at all in Access,
since as far as I know it is not an Access function. DateDiff is the usual
way of calculating ages. The following, which borrows from some newsgroup
postings and from the Date/Time section of http://www.mvps.org/access/,
seems to work by calculating the number of days and dividing by 365.25.
There may be some minor imprecision in some cases, but you can decide if
that matters.
AgeAtDeath:
IIf(DateDiff("yyyy",[DoB],[DoD])=0,"",Format(DateDiff("d",[DoB],[DoD])/365.25,"Fixed"))
Simplest way to sort and filter may be to add another field to the query:
DateCriteria: DateDiff("yyyy",[DoB],[DoD])
You can sort and filter on that field. By the way, "Fixed" seems to be the
way the format "0.00" ends up. Same thing. Also, you can use Date()
instead of [DoD] to calculate a living person's age.
Access serves a very different purpose than Excel. If you have multiple
incidents for a single patient, is your approach to enter another complete
record? If so, you are taking the long way around to avoid using the very
aspect of Access that makes it a useful. The best way to do what you need
is to have a Patients table with a record for each patient, and a related
Incidents table to keep track of each individual incident. The only real
way of entering multiple incidents for one patient in Excel is to enter the
patient information over and over. In Access you enter the patient
information once, and relate other information to that record.
SQL is (among other things) the language behind a query. You can choose SQL
View for a query to see how that looks. You can either enter the SQL
directly, or you can use the design view to build the SQL statement.
Parameter queries are the way Access provides for filtering data based on a
particular field.
I suspect you are frustrated because an Access table resembles a
spreadsheet, but it is very different. Both have their uses, but they are
not interchangeable. Your efforts seem to be centered on making an Access
table behave like an Excel spreadsheet. It isn't going to happen, but maybe
the folks in this group could help you get the most out of Access by making
use of its capabilities.

"Myrna Larson" wrote in message
...
This is an MDB database, not SQL. There is only 1 table (it's really just
a
small flat-file which shouldn't require a relational DB, but see comment
below). There are just 700 patients.

Some persons have a genetically determined disease, others do not. Of the
affected persons, some have been treated with surgery, others have not. So
basically there are 3 groups: unaffected, affected but not treated, and
affected and treated.

We are tracking longevity and the occurrence and timing of heart attacks
and
strokes in the 3 groups. I have created all of the queries using the QBE
grid;
I haven't needed parameter queries. The form in question is used primarily
for
data entry but on occasion I want to filter by age as well.

Each record contains data about a single patient, about 32 fields. There
are 5
date/time fields: date of birth, date of surgery, date of heart attack or
stroke, date of last contact, and date of death. Any or all of these
fields
can be null. I calculate 4 "ages": age at surgery, age at heart attack,
age at
last contact, and age at death. For a living patient with no surgery and
no
heart attack, only age at last contact is not null.

This data could easily be tracked as a flat-file in Excel, except for 3
Excel
problems: Excel can't handle dates prior to 1/1/1900 (some of these
patients
were born in the 1800's), the user can fairly easily "scramble" the table
by
improper sorting, and it's much more difficult to set up the "queries" and
filtering in Excel.

BUT... in Excel, I can easily create a data entry form and specify a
number
format that (1) does not display zero values, (2) ALWAYS shows 2 decimal
places for non-zero values, and (3) achieving that format doesn't require
changing the type of the underlying data to text.

I don't seem to be able to do that in Access.

This is so EASY in Excel; I don't know why it should be so DIFFICULT
(impossible?) in Access. But I guess that's the way it is. Maybe the next
version of Excel will allow dates prior to 1/1/1900.

Anyway, thank you for your help.

PS: I have from a 25 year background in Basic programming and use of
Excel. (I
was a VB and Excel MVP from 1994-2004.) But I am essentially a newbie when
it
comes to Access.


On 18 Nov 2005 01:08:00 -0800, wrote:


Myrna Larson wrote:
Are you suggesting that I should not use the built-in Filter-by-Form to
allow
the user to set the filter criteria, but that I should develop my own
form
for
that purpose?

One step at a time: you need the data layer elements before you can
build the front end elements (didn't you say the built in features
weren't working for you?)

You've given no details about your base tables, constraints, business
rules for age calculations, etc so I'll fake it with this test table:

CREATE TABLE Test (
key_col INTEGER NOT NULL UNIQUE,
DoB DATETIME NOT NULL,
DoD DATETIME
)
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (1, CDATE('1970-01-01'), CDATE('2005-11-01'))
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (2, CDATE('1940-02-02'), CDATE('2005-02-01'))
;
INSERT INTO Test (key_col, DoB, DoD)
VALUES (3, CDATE('1980-03-03'), NULL)
;

I'll fake the calculation to suit, too. My formatting should be as per
your spec, though. If you are not using 'ANSI SQL-92 query mode' or
ADO, you may find it quicker to paste the following SELECT into a new
Query object and save it manually:

CREATE VIEW TestView
AS
SELECT key_col,
IIF(DoD IS NULL, 0, DATEDIFF("d", DoB, DoD) / 365.25) AS age_at_death,
IIF(age_at_death = 0,
'',
FORMAT$(age_at_death,
IIF(RIGHT$(FORMAT$(age_at_death, '##0.00'), 2) = '00',
'##0',
'##0.00'
)
)
) AS display_age_at_death
FROM Test
;

Now we need another element in the data layer which will take a
strongly typed parameter for age (I assume you want to query on whole
years) and the applicable operators can be encoded. Again, if you want
to do this manually you can use a SELECT in a Query object using the
PARAMETERS declaration but you will have to forego the default
parameter value:

CREATE PROCEDURE TestProc (
:value INTEGER,
perator_code INTEGER = 1
) AS
SELECT key_col, age_at_death, display_age_at_death
FROM TestView
WHERE SWITCH(
age_at_death = 0, 0,
perator_code = 1, IIF(age_at_death = :value, 1, 0),
perator_code = 2, IIF(age_at_death :value, 1, 0),
perator_code = 3, IIF(age_at_death :value, 1, 0),
perator_code = 4, IIF(age_at_death = :value, 1, 0),
perator_code = 5, IIF(age_at_death = :value, 1, 0),
perator_code = 6, IIF(age_at_death :value, 1, 0),
TRUE, 0) = 1
;

Now, to query on 40, use

EXEC TestProc 40, 3

With the required data layer elements in place, you can collect the
parameter values using forms and use them to call the proc/Parameter
Query.

I'll add my comments in-line below.

I'm not sure how (or if) you got the Age function to work at all in Access,
since as far as I know it is not an Access function.

A query in an MDB file CAN call a user-defined (i.e. VBA) function. I have
been doing that since I started this project, using Access 2000. I've now
upgraded to Access 2003.

How did I get it to work? The VBA code is in a module in the MDB file. The
function name is Age and it returns a double. I created the query in Design
View, using the grid. e.g. to calculate age at death, I use fields named DoB
and DoD. In the Field row of the query, I wrote AgeAtDeath:Age(DoB,DoD).

calculating the number of days and dividing by 365.25.
There may be some minor imprecision in some cases, but you can decide if
that matters.

AgeAtDeath:
IIf(DateDiff("yyyy",[DoB],[DoD])=0,"",Format(DateDiff("d",[DoB],[DoD])/365.25,"Fixed"))
Simplest way to sort and filter may be to add another field to the query:
DateCriteria: DateDiff("yyyy",[DoB],[DoD])

I guess you haven't read any of my messages in the Excel forums in response to
formulas using 365.25 as the number of days in a year. Actually they may be
better described as "harangues" than "responses" g.

That formula is much more complicated than Age(Dob,Dod), I think you'll agree.
And for a patient who was born in, say, 1 Nov 2004, and died on 1 Feb 2005,
your first DateDiff will give a result of 0. I don't want 0, I want 0.25. And
there will be errors when the 2nd date is very close to the anniversary of the
birth date. The user will certainly "notice" if a patient was born on Jan 15,
1940, a heart attack occurred on Jan 15, 1989, and the age is shown as 48.99
or 49.01.

Also, you can use Date()
instead of [DoD] to calculate a living person's age.

No, I can't. I must use the date on which we last contacted the patient.
He/she could have died or had a heart attack, etc, since the last contact.

Access serves a very different purpose than Excel. If you have multiple
incidents for a single patient, is your approach to enter another complete
record? If so, you are taking the long way around to avoid using the very
aspect of Access that makes it a useful. The best way to do what you need
is to have a Patients table with a record for each patient, and a related
Incidents table to keep track of each individual incident. The only real
way of entering multiple incidents for one patient in Excel is to enter the
patient information over and over. In Access you enter the patient
information once, and relate other information to that record.

I do understand the concept of a normalized data structure. I believe I have
it in my file. There is no repeated information from one record to the next.

I am only interested in the date and type of the FIRST incident. The patient
is removed from the study at that point. We are, however keeping track of the
count of heart attacks, I have a field for the date of the 1st event, what it
was (heart attack, stroke, etc), plus 4 numeric fields for the counts of the 4
kinds of problems we are tracking (heart attack, stroke, coronary artery and
carotid artery surgery).

I mentioned in an earlier post the problems with Excel that require me to use
Access.

SQL is (among other things) the language behind a query.

Yes, I know that, and I have used it. When editing an SQL statement, I often
paste it into my editor program, where search and replace is available. In
another application, I use stock prices stored in an MDB file. I import those
into a spreadsheet using SQL queries that I contruct in the VBA code in the
spreadsheet. So I am quite familiar with SQL, though I wouldn't say I'm an
expert.

Parameter queries are the way Access provides for filtering data based on a
particular field.

Is that what is happening "behind the scenes" when you use Filter-by-Form with
a Form? Or is the Filter property of the record set being changed?

I suspect you are frustrated because an Access table resembles a
spreadsheet, but it is very different. Both have their uses, but they are
not interchangeable.

No, that's not the issue. I understand the difference between a database and a
spreadsheet, and when to use which.

My question pertained to problems with a FORM, not a datasheet, and that form
doesn't resemble a spreadsheet in any way. (I need a form for data entry: I
have too many fields to do it in datasheet view, one of them is a memo field
of several hundred characters, including new lines, so editing them from
Datasheet view is not really an option. In addition, I have lots of error
checking in my BeforeUpdate event procedure.)

Your efforts seem to be centered on making an Access
table behave like an Excel spreadsheet.

Not at all! I just want control over the *display format* of a number that's
shown in a text box on a form. I've read messages here that say you can't
really apply formatting to a datasheet view of a table or query -- that has to
be done in the form or report. And that's exactly what I have been trying to
accomplish: the format on the FORM.

I always want 2 decimal places, or when the result is 0, an "apparently" blank
cell. And I want to accomplish this WITHOUT changing the date type of the
underlying data from numeric to text or variant, so I can still filter on this
as a numeric field. And there IS a way to do all of that without an additional
field.

I guess you didn't see my last post, where I said I had solved the problem.
The answer is in the Format property for the field.

The problem I was having was that I entered this as the format:

0.00;;;

and Access changed that to a specification with only 1 section, i.e.

Fixed

My fundamental problem was not being able to find Help for "custom number
formats". Even though Help uses that term, the file is not indexed on it. You
have to search for help on "Format Property". The 2nd problem is that the
number format behaves a bit differently in Access than Excel, and without
Help, I didn't know that. The solution was to type the format string as

0.00; ; ; ;

Quite straightforward, once you know the "trick".

It isn't going to happen, but maybe
the folks in this group could help you get the most out of Access by making
use of its capabilities.

I think I am getting what I need from Access now that this latest irritation
has been resolved.

Thanks anyway for your thoughts.