Average with Time intervals

[Susan]

Hello,

I posted (creating relationships with time series) yesterday within "access
Database Design".

Basically I have 24 tables that are linked to txt files (the output of my
instrument). Yesterdays question concerned how to create the relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different cells. I usually
run experiments where I duplicate "treatments" in 3 or more cells.

My data base contains the linked txt files, a "details" table which contains
the columns "cell, treatment, start time"; and the query created yesterday
which collates all of "hour" (time elapsed) and "mL gas" and cell into one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but with averages of
mLgas, based on "treatment". For example cells 3, 15 and 24 are one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same time, I would
like to be able to average them on a half hour basis. My current crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell was started
slightly different times, the result of the average doesn't make sense; all
values should have been included in one row.

Can anyone help please???

Susan

 

[Jeff Boyce]

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How does Access
know that? Unless you (or Access) has a way to connect those together as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design that
(apparently) uses one table per ?cell. If this is true, this is not
particularly normalized, and will make your analysis much harder.

 

[Susan]

Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but that's how
I have to do it. I have to link to the txt files as outputted from the
instrument. I can not change the format of these files and linking to them
allows me to keep them up to date (the instrument writes to the txt file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be able to do it
with Time instead (and these become out of sync). Also, occassionally the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half hour
intervals but I can be flexiable on this.

Cheers

Susan

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How does Access
know that? Unless you (or Access) has a way to connect those together as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design that
(apparently) uses one table per ?cell. If this is true, this is not
particularly normalized, and will make your analysis much harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday within "access
Database Design".

Basically I have 24 tables that are linked to txt files (the output of my
instrument). Yesterdays question concerned how to create the relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different cells. I usually
run experiments where I duplicate "treatments" in 3 or more cells.

My data base contains the linked txt files, a "details" table which contains
the columns "cell, treatment, start time"; and the query created yesterday
which collates all of "hour" (time elapsed) and "mL gas" and cell into one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but with averages of
mLgas, based on "treatment". For example cells 3, 15 and 24 are one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same time, I would
like to be able to average them on a half hour basis. My current crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell was started
slightly different times, the result of the average doesn't make sense; all
values should have been included in one row.

Can anyone help please???

Susan

 

[Jeff Boyce]

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not getting the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use text files.
Another approach is to parse the data into a well-normalized table using
queries...

Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but that's how
I have to do it. I have to link to the txt files as outputted from the
instrument. I can not change the format of these files and linking to them
allows me to keep them up to date (the instrument writes to the txt file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be able to do it
with Time instead (and these become out of sync). Also, occassionally the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half hour
intervals but I can be flexiable on this.

Cheers

Susan

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How does Access
know that? Unless you (or Access) has a way to connect those together as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design that
(apparently) uses one table per ?cell. If this is true, this is not
particularly normalized, and will make your analysis much harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday within "access
Database Design".

Basically I have 24 tables that are linked to txt files (the output of my
instrument). Yesterdays question concerned how to create the relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different cells. I usually
run experiments where I duplicate "treatments" in 3 or more cells.

My data base contains the linked txt files, a "details" table which contains
the columns "cell, treatment, start time"; and the query created yesterday
which collates all of "hour" (time elapsed) and "mL gas" and cell into one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but with

averages
of

mLgas, based on "treatment". For example cells 3, 15 and 24 are one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same time, I would
like to be able to average them on a half hour basis. My current crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell was started
slightly different times, the result of the average doesn't make

sense;
all

values should have been included in one row.

Can anyone help please???

Susan

 

[Susan]

Hello Jeff,

Totally understand the concept of under-caffeinated. I'm doing Home Reno's
in the evenings and I'm low on my quota of sleep.

The problem is that each cell is seperated by a couple of minutes. When I
average the mL gas, it does it at each "time". What I would like to be able
to do is average the mL gas for each treatment (multiple cells) in hour
blocks; say between 25/11/05 11:31 to 12:00, then between 12:01 and to 12:30
etc.

In the table below (the results of a cross-tab query), "Manifold" is the
treatment. The three results (1.23, 0, 2.8) are the three cell results. This
crosstab query averaging the results...but because each cell is 1 minute
seperate from each other it's not producing the number I'm after (1.35).

When I plot these on a scatter plot ~11:00 plots three dots (1.26, 0, 2.8).
I need to plot them as one point (1.35)

Time Manifold
25/11/2005 11:00:00 AM 1.26
25/11/2005 11:01:48 AM 0
25/11/2005 11:02:24 AM 2.8

I hope this is making more sense. I guess I should have said averaging mL of
gas in a "block/ interval" of time.

My middle of the night thought last night was to write a VERY long iif
statement...but thsi is going to HORRIBLE as these experiments can run for
several months.

Thanks again

Susan

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not getting the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use text files.
Another approach is to parse the data into a well-normalized table using
queries...

Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but that's how
I have to do it. I have to link to the txt files as outputted from the
instrument. I can not change the format of these files and linking to them
allows me to keep them up to date (the instrument writes to the txt file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be able to do it
with Time instead (and these become out of sync). Also, occassionally the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half hour
intervals but I can be flexiable on this.

Cheers

Susan

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How does Access
know that? Unless you (or Access) has a way to connect those together as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design that
(apparently) uses one table per ?cell. If this is true, this is not
particularly normalized, and will make your analysis much harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday within
"access
Database Design".

Basically I have 24 tables that are linked to txt files (the output of my
instrument). Yesterdays question concerned how to create the relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different cells. I
usually
run experiments where I duplicate "treatments" in 3 or more cells.

My data base contains the linked txt files, a "details" table which
contains
the columns "cell, treatment, start time"; and the query created yesterday
which collates all of "hour" (time elapsed) and "mL gas" and cell into one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but with averages
of
mLgas, based on "treatment". For example cells 3, 15 and 24 are one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same time, I would
like to be able to average them on a half hour basis. My current crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell was started
slightly different times, the result of the average doesn't make sense;
all
values should have been included in one row.

Can anyone help please???

Susan

 

[Jeff Boyce]

Susan

I don't have the same excuse tonight, but I'm still fuzzy on the concept.

If you are saying you have a large set of data, could you write a procedure
that does the calculation? This would save you the multiple embedded IIF()
statement.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff,

Totally understand the concept of under-caffeinated. I'm doing Home Reno's
in the evenings and I'm low on my quota of sleep.

The problem is that each cell is seperated by a couple of minutes. When I
average the mL gas, it does it at each "time". What I would like to be able
to do is average the mL gas for each treatment (multiple cells) in hour
blocks; say between 25/11/05 11:31 to 12:00, then between 12:01 and to 12:30
etc.

In the table below (the results of a cross-tab query), "Manifold" is the
treatment. The three results (1.23, 0, 2.8) are the three cell results. This
crosstab query averaging the results...but because each cell is 1 minute
seperate from each other it's not producing the number I'm after (1.35).

When I plot these on a scatter plot ~11:00 plots three dots (1.26, 0, 2.8).
I need to plot them as one point (1.35)

Time Manifold
25/11/2005 11:00:00 AM 1.26
25/11/2005 11:01:48 AM 0
25/11/2005 11:02:24 AM 2.8

I hope this is making more sense. I guess I should have said averaging mL of
gas in a "block/ interval" of time.

My middle of the night thought last night was to write a VERY long iif
statement...but thsi is going to HORRIBLE as these experiments can run for
several months.

Thanks again

Susan

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not getting the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use text files.
Another approach is to parse the data into a well-normalized table using
queries...

Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but

that's
how

I have to do it. I have to link to the txt files as outputted from the
instrument. I can not change the format of these files and linking to them
allows me to keep them up to date (the instrument writes to the txt file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be able to

do
it

with Time instead (and these become out of sync). Also, occassionally the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half hour
intervals but I can be flexiable on this.

Cheers

Susan

:

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How does Access
know that? Unless you (or Access) has a way to connect those

together
as

one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design that
(apparently) uses one table per ?cell. If this is true, this is not
particularly normalized, and will make your analysis much harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday within
"access
Database Design".

Basically I have 24 tables that are linked to txt files (the

output of
my

instrument). Yesterdays question concerned how to create the relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different cells. I
usually
run experiments where I duplicate "treatments" in 3 or more cells.

My data base contains the linked txt files, a "details" table which
contains
the columns "cell, treatment, start time"; and the query created yesterday
which collates all of "hour" (time elapsed) and "mL gas" and cell

into
one

query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but with averages
of
mLgas, based on "treatment". For example cells 3, 15 and 24 are one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same time, I would
like to be able to average them on a half hour basis. My current crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell was started
slightly different times, the result of the average doesn't make sense;
all
values should have been included in one row.

Can anyone help please???

Susan

 

[Susan]

Hey Jeff,

I'm not sure how else to explain my problem! Without showing you the graph
that access produces vrs the graph that I want to produce

I had a go at writing an iif statement. It came out looking this (and sorry
this is a real mess):

Time_Interval: IIf([time]<([minOfStart time]+(6/24)),[minOfStart
time],(IIf([time]<([minOfStart time]+(12/24)),([minOfStart
time]+(6/24)),(IIf([time]<([minOfStart time]+(18/24)),([minOfStart
time]+(12/24)),(IIf([time]<([minOfStart time]+(1)),([minOfStart
time]+(18/24)),(IIf([time]<([minOfStart time]+(1+6/24)),([minOfStart
time]+(1+12/24)),(IIf([time]<([minOfStart time]+(1+12/24)),([minOfStart
time]+(1+6/24)),(IIf([time]<([minOfStart time]+(1+18/24)),([minOfStart
time]+(1+12/24)),(IIf([time]<([minOfStart time]+(2)),([minOfStart
time]+(1+18/24)),-1)))))))))))))))

Where "minOfStart time" is the first cell to start (the minimium (smallest)
start time).
Time is "start time"+([hour]/24)

This works for the 2 days (if I break the days into 6 hour "intervals"), but
I can't embed any more iif statements - access has a wobbly and tells me "too
complicated"; I would prefer to use 1 hour (or even 30 minute intervals).

I know that there is the INT function in access - but don't know how to use
it.

I also don't know about "proceedures"

Thanks again.

Susan

Susan

I don't have the same excuse tonight, but I'm still fuzzy on the concept.

If you are saying you have a large set of data, could you write a procedure
that does the calculation? This would save you the multiple embedded IIF()
statement.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff,

Totally understand the concept of under-caffeinated. I'm doing Home Reno's
in the evenings and I'm low on my quota of sleep.

The problem is that each cell is seperated by a couple of minutes. When I
average the mL gas, it does it at each "time". What I would like to be able
to do is average the mL gas for each treatment (multiple cells) in hour
blocks; say between 25/11/05 11:31 to 12:00, then between 12:01 and to 12:30
etc.

In the table below (the results of a cross-tab query), "Manifold" is the
treatment. The three results (1.23, 0, 2.8) are the three cell results. This
crosstab query averaging the results...but because each cell is 1 minute
seperate from each other it's not producing the number I'm after (1.35).

When I plot these on a scatter plot ~11:00 plots three dots (1.26, 0, 2.8).
I need to plot them as one point (1.35)

Time Manifold
25/11/2005 11:00:00 AM 1.26
25/11/2005 11:01:48 AM 0
25/11/2005 11:02:24 AM 2.8

I hope this is making more sense. I guess I should have said averaging mL of
gas in a "block/ interval" of time.

My middle of the night thought last night was to write a VERY long iif
statement...but thsi is going to HORRIBLE as these experiments can run for
several months.

Thanks again

Susan

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not getting the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use text files.
Another approach is to parse the data into a well-normalized table using
queries...

Regards

Jeff Boyce
<Office/Access MVP>


Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but that's
how
I have to do it. I have to link to the txt files as outputted from the
instrument. I can not change the format of these files and linking to them
allows me to keep them up to date (the instrument writes to the txt file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be able to do
it
with Time instead (and these become out of sync). Also, occassionally the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half hour
intervals but I can be flexiable on this.

Cheers

Susan

:

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How does
Access
know that? Unless you (or Access) has a way to connect those together
as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design that
(apparently) uses one table per ?cell. If this is true, this is not
particularly normalized, and will make your analysis much harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday within
"access
Database Design".

Basically I have 24 tables that are linked to txt files (the output of
my
instrument). Yesterdays question concerned how to create the
relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different cells. I
usually
run experiments where I duplicate "treatments" in 3 or more cells.

My data base contains the linked txt files, a "details" table which
contains
the columns "cell, treatment, start time"; and the query created
yesterday
which collates all of "hour" (time elapsed) and "mL gas" and cell into
one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but with
averages
of
mLgas, based on "treatment". For example cells 3, 15 and 24 are one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same time, I
would
like to be able to average them on a half hour basis. My current
crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell was
started
slightly different times, the result of the average doesn't make
sense;
all
values should have been included in one row.

Can anyone help please???

Susan

 

[John Spencer]

Pardon me, I may be offbase here.

Susan,
Do you mean you want to group the readings into 30-minute intervals and then
get an average for each of the 30-minute intervals?

Sincerely,
John

Susan

I don't have the same excuse tonight, but I'm still fuzzy on the concept.

If you are saying you have a large set of data, could you write a
procedure
that does the calculation? This would save you the multiple embedded
IIF()
statement.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff,

Totally understand the concept of under-caffeinated. I'm doing Home
Reno's
in the evenings and I'm low on my quota of sleep.

The problem is that each cell is seperated by a couple of minutes. When I
average the mL gas, it does it at each "time". What I would like to be able
to do is average the mL gas for each treatment (multiple cells) in hour
blocks; say between 25/11/05 11:31 to 12:00, then between 12:01 and to 12:30
etc.

In the table below (the results of a cross-tab query), "Manifold" is the
treatment. The three results (1.23, 0, 2.8) are the three cell results. This
crosstab query averaging the results...but because each cell is 1 minute
seperate from each other it's not producing the number I'm after (1.35).

When I plot these on a scatter plot ~11:00 plots three dots (1.26, 0, 2.8).
I need to plot them as one point (1.35)

Time Manifold
25/11/2005 11:00:00 AM 1.26
25/11/2005 11:01:48 AM 0
25/11/2005 11:02:24 AM 2.8

I hope this is making more sense. I guess I should have said averaging mL of
gas in a "block/ interval" of time.

My middle of the night thought last night was to write a VERY long iif
statement...but thsi is going to HORRIBLE as these experiments can run
for
several months.

Thanks again

Susan

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not getting the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use text files.
Another approach is to parse the data into a well-normalized table
using
queries...

Regards

Jeff Boyce
<Office/Access MVP>


Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but that's
how
I have to do it. I have to link to the txt files as outputted from
the
instrument. I can not change the format of these files and linking to them
allows me to keep them up to date (the instrument writes to the txt file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be able
to do
it
with Time instead (and these become out of sync). Also, occassionally the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half hour
intervals but I can be flexiable on this.

Cheers

Susan

:

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How
does
Access
know that? Unless you (or Access) has a way to connect those together
as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design
that
(apparently) uses one table per ?cell. If this is true, this is
not
particularly normalized, and will make your analysis much harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday within
"access
Database Design".

Basically I have 24 tables that are linked to txt files (the output of
my
instrument). Yesterdays question concerned how to create the
relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different cells. I
usually
run experiments where I duplicate "treatments" in 3 or more
cells.

My data base contains the linked txt files, a "details" table which
contains
the columns "cell, treatment, start time"; and the query created
yesterday
which collates all of "hour" (time elapsed) and "mL gas" and cell into
one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but with
averages
of
mLgas, based on "treatment". For example cells 3, 15 and 24 are one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same time,
I
would
like to be able to average them on a half hour basis. My current
crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell
was
started
slightly different times, the result of the average doesn't make
sense;
all
values should have been included in one row.

Can anyone help please???

Susan

 

[Susan]

Precisely! That's exactly what I want to be able to do.

Thanks

Pardon me, I may be offbase here.

Susan,
Do you mean you want to group the readings into 30-minute intervals and then
get an average for each of the 30-minute intervals?

Sincerely,
John

Susan

I don't have the same excuse tonight, but I'm still fuzzy on the concept.

If you are saying you have a large set of data, could you write a
procedure
that does the calculation? This would save you the multiple embedded
IIF()
statement.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff,

Totally understand the concept of under-caffeinated. I'm doing Home
Reno's
in the evenings and I'm low on my quota of sleep.

The problem is that each cell is seperated by a couple of minutes. When I
average the mL gas, it does it at each "time". What I would like to be able
to do is average the mL gas for each treatment (multiple cells) in hour
blocks; say between 25/11/05 11:31 to 12:00, then between 12:01 and to 12:30
etc.

In the table below (the results of a cross-tab query), "Manifold" is the
treatment. The three results (1.23, 0, 2.8) are the three cell results. This
crosstab query averaging the results...but because each cell is 1 minute
seperate from each other it's not producing the number I'm after (1.35).

When I plot these on a scatter plot ~11:00 plots three dots (1.26, 0, 2.8).
I need to plot them as one point (1.35)

Time Manifold
25/11/2005 11:00:00 AM 1.26
25/11/2005 11:01:48 AM 0
25/11/2005 11:02:24 AM 2.8

I hope this is making more sense. I guess I should have said averaging mL of
gas in a "block/ interval" of time.

My middle of the night thought last night was to write a VERY long iif
statement...but thsi is going to HORRIBLE as these experiments can run
for
several months.

Thanks again

Susan


:

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not getting the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use text files.
Another approach is to parse the data into a well-normalized table
using
queries...

Regards

Jeff Boyce
<Office/Access MVP>


Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but that's
how
I have to do it. I have to link to the txt files as outputted from
the
instrument. I can not change the format of these files and linking to them
allows me to keep them up to date (the instrument writes to the txt file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be able
to do
it
with Time instead (and these become out of sync). Also, occassionally the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half hour
intervals but I can be flexiable on this.

Cheers

Susan

:

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How
does
Access
know that? Unless you (or Access) has a way to connect those together
as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design
that
(apparently) uses one table per ?cell. If this is true, this is
not
particularly normalized, and will make your analysis much harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday within
"access
Database Design".

Basically I have 24 tables that are linked to txt files (the output of
my
instrument). Yesterdays question concerned how to create the
relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different cells. I
usually
run experiments where I duplicate "treatments" in 3 or more
cells.

My data base contains the linked txt files, a "details" table which
contains
the columns "cell, treatment, start time"; and the query created
yesterday
which collates all of "hour" (time elapsed) and "mL gas" and cell into
one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but with
averages
of
mLgas, based on "treatment". For example cells 3, 15 and 24 are one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same time,
I
would
like to be able to average them on a half hour basis. My current
crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell
was
started
slightly different times, the result of the average doesn't make
sense;
all
values should have been included in one row.

Can anyone help please???

Susan

 

[John Spencer]

To "round" your date to 30 minute intervals, you can use

DateAdd("n",30*(DateDiff("n",0,[Your Date Field])\30),CDate(0))

There may be more efficient ways to do this, but this is what I've come up
with.

Precisely! That's exactly what I want to be able to do.

Thanks

Pardon me, I may be offbase here.

Susan,
Do you mean you want to group the readings into 30-minute intervals and
then
get an average for each of the 30-minute intervals?

Sincerely,
John

message

Susan

I don't have the same excuse tonight, but I'm still fuzzy on the
concept.

If you are saying you have a large set of data, could you write a
procedure
that does the calculation? This would save you the multiple embedded
IIF()
statement.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff,

Totally understand the concept of under-caffeinated. I'm doing Home
Reno's
in the evenings and I'm low on my quota of sleep.

The problem is that each cell is seperated by a couple of minutes.
When I
average the mL gas, it does it at each "time". What I would like to be
able
to do is average the mL gas for each treatment (multiple cells) in
hour
blocks; say between 25/11/05 11:31 to 12:00, then between 12:01 and to
12:30
etc.

In the table below (the results of a cross-tab query), "Manifold" is
the
treatment. The three results (1.23, 0, 2.8) are the three cell
results.
This
crosstab query averaging the results...but because each cell is 1
minute
seperate from each other it's not producing the number I'm after
(1.35).

When I plot these on a scatter plot ~11:00 plots three dots (1.26, 0,
2.8).
I need to plot them as one point (1.35)

Time Manifold
25/11/2005 11:00:00 AM 1.26
25/11/2005 11:01:48 AM 0
25/11/2005 11:02:24 AM 2.8

I hope this is making more sense. I guess I should have said averaging
mL
of
gas in a "block/ interval" of time.

My middle of the night thought last night was to write a VERY long iif
statement...but thsi is going to HORRIBLE as these experiments can run
for
several months.

Thanks again

Susan


:

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not getting
the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use
text
files.
Another approach is to parse the data into a well-normalized table
using
queries...

Regards

Jeff Boyce
<Office/Access MVP>


Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but
that's
how
I have to do it. I have to link to the txt files as outputted from
the
instrument. I can not change the format of these files and linking
to
them
allows me to keep them up to date (the instrument writes to the
txt
file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be
able
to
do
it
with Time instead (and these become out of sync). Also,
occassionally
the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half
hour
intervals but I can be flexiable on this.

Cheers

Susan

:

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How
does
Access
know that? Unless you (or Access) has a way to connect those
together
as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design
that
(apparently) uses one table per ?cell. If this is true, this is
not
particularly normalized, and will make your analysis much
harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday
within
"access
Database Design".

Basically I have 24 tables that are linked to txt files (the
output of
my
instrument). Yesterdays question concerned how to create the
relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different
cells.
I
usually
run experiments where I duplicate "treatments" in 3 or more
cells.

My data base contains the linked txt files, a "details" table
which
contains
the columns "cell, treatment, start time"; and the query
created
yesterday
which collates all of "hour" (time elapsed) and "mL gas" and
cell
into
one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but
with
averages
of
mLgas, based on "treatment". For example cells 3, 15 and 24
are
one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same
time,
I
would
like to be able to average them on a half hour basis. My
current
crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell
was
started
slightly different times, the result of the average doesn't
make
sense;
all
values should have been included in one row.

Can anyone help please???

Susan

 

[Susan]

Hello John,

Perfect! Wonderful - thanks so much. It works a treat. I don't understand
the formula but it works

Cheers

Susan

To "round" your date to 30 minute intervals, you can use

DateAdd("n",30*(DateDiff("n",0,[Your Date Field])\30),CDate(0))

There may be more efficient ways to do this, but this is what I've come up
with.

Precisely! That's exactly what I want to be able to do.

Thanks

Pardon me, I may be offbase here.

Susan,
Do you mean you want to group the readings into 30-minute intervals and
then
get an average for each of the 30-minute intervals?

Sincerely,
John

message
Susan

I don't have the same excuse tonight, but I'm still fuzzy on the
concept.

If you are saying you have a large set of data, could you write a
procedure
that does the calculation? This would save you the multiple embedded
IIF()
statement.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff,

Totally understand the concept of under-caffeinated. I'm doing Home
Reno's
in the evenings and I'm low on my quota of sleep.

The problem is that each cell is seperated by a couple of minutes.
When I
average the mL gas, it does it at each "time". What I would like to be
able
to do is average the mL gas for each treatment (multiple cells) in
hour
blocks; say between 25/11/05 11:31 to 12:00, then between 12:01 and to
12:30
etc.

In the table below (the results of a cross-tab query), "Manifold" is
the
treatment. The three results (1.23, 0, 2.8) are the three cell
results.
This
crosstab query averaging the results...but because each cell is 1
minute
seperate from each other it's not producing the number I'm after
(1.35).

When I plot these on a scatter plot ~11:00 plots three dots (1.26, 0,
2.8).
I need to plot them as one point (1.35)

Time Manifold
25/11/2005 11:00:00 AM 1.26
25/11/2005 11:01:48 AM 0
25/11/2005 11:02:24 AM 2.8

I hope this is making more sense. I guess I should have said averaging
mL
of
gas in a "block/ interval" of time.

My middle of the night thought last night was to write a VERY long iif
statement...but thsi is going to HORRIBLE as these experiments can run
for
several months.

Thanks again

Susan


:

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not getting
the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use
text
files.
Another approach is to parse the data into a well-normalized table
using
queries...

Regards

Jeff Boyce
<Office/Access MVP>


Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great - but
that's
how
I have to do it. I have to link to the txt files as outputted from
the
instrument. I can not change the format of these files and linking
to
them
allows me to keep them up to date (the instrument writes to the
txt
file
every half hour.

I have a union query which collates all the individual txt files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be
able
to
do
it
with Time instead (and these become out of sync). Also,
occassionally
the
instrument hiccups and hours also become slightly out of sync too.

I need to be able to average mL in time intervals, ideally half
hour
intervals but I can be flexiable on this.

Cheers

Susan

:

Susan

How do you know that cells 3, 15 & 24 are one "treatment"? How
does
Access
know that? Unless you (or Access) has a way to connect those
together
as
one treatment, I don't see how you can average their durations.

I didn't see your previous post, but wonder about a data design
that
(apparently) uses one table per ?cell. If this is true, this is
not
particularly normalized, and will make your analysis much
harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series) yesterday
within
"access
Database Design".

Basically I have 24 tables that are linked to txt files (the
output of
my
instrument). Yesterdays question concerned how to create the
relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different
cells.
I
usually
run experiments where I duplicate "treatments" in 3 or more
cells.

My data base contains the linked txt files, a "details" table
which
contains
the columns "cell, treatment, start time"; and the query
created
yesterday
which collates all of "hour" (time elapsed) and "mL gas" and
cell
into
one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but
with
averages
of
mLgas, based on "treatment". For example cells 3, 15 and 24
are
one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same
time,
I
would
like to be able to average them on a half hour basis. My
current
crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each cell
was
started
slightly different times, the result of the average doesn't
make
sense;
all
values should have been included in one row.

Can anyone help please???

Susan

 

[John Spencer]

DateAdd("n",30*(DateDiff("n",0,[Your Date Field])\30),CDate(0))
Explanation:
DateDiff("n",0,[Your Date Field]) : This gets the number of minutes since
the zero date (You could actually use any date here, such as #1/1/1950#)

\30 does an integer divide of the number of minutes, so there is no
fractional remainder.
Then we multiply by 30 (in effect we have rounded the minutes down to the
nearest 30.

DateAdd("n",<<Our minutes calculation>>, CDate(0)) : Adds the number of
minutes back to the zero date. CDate(0) could again be replaced with any
date (as long as it is the same one used earlier in the calculation -
#1/1/1950#).

So using Jan 1 1950 the formula would look like:
DateAdd("n",30*(DateDiff("n",#1/1/1950#,[Your Date Field])\30),#1/1/1950#)

I hope that helps you understand it.

Hello John,

Perfect! Wonderful - thanks so much. It works a treat. I don't understand
the formula but it works

Cheers

Susan

To "round" your date to 30 minute intervals, you can use

DateAdd("n",30*(DateDiff("n",0,[Your Date Field])\30),CDate(0))

There may be more efficient ways to do this, but this is what I've come
up
with.

Precisely! That's exactly what I want to be able to do.

Thanks

:

Pardon me, I may be offbase here.

Susan,
Do you mean you want to group the readings into 30-minute intervals
and
then
get an average for each of the 30-minute intervals?

Sincerely,
John

message
Susan

I don't have the same excuse tonight, but I'm still fuzzy on the
concept.

If you are saying you have a large set of data, could you write a
procedure
that does the calculation? This would save you the multiple
embedded
IIF()
statement.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello Jeff,

Totally understand the concept of under-caffeinated. I'm doing Home
Reno's
in the evenings and I'm low on my quota of sleep.

The problem is that each cell is seperated by a couple of minutes.
When I
average the mL gas, it does it at each "time". What I would like to
be
able
to do is average the mL gas for each treatment (multiple cells) in
hour
blocks; say between 25/11/05 11:31 to 12:00, then between 12:01 and
to
12:30
etc.

In the table below (the results of a cross-tab query), "Manifold"
is
the
treatment. The three results (1.23, 0, 2.8) are the three cell
results.
This
crosstab query averaging the results...but because each cell is 1
minute
seperate from each other it's not producing the number I'm after
(1.35).

When I plot these on a scatter plot ~11:00 plots three dots (1.26,
0,
2.8).
I need to plot them as one point (1.35)

Time Manifold
25/11/2005 11:00:00 AM 1.26
25/11/2005 11:01:48 AM 0
25/11/2005 11:02:24 AM 2.8

I hope this is making more sense. I guess I should have said
averaging
mL
of
gas in a "block/ interval" of time.

My middle of the night thought last night was to write a VERY long
iif
statement...but thsi is going to HORRIBLE as these experiments can
run
for
several months.

Thanks again

Susan


:

Susan

Perhaps I'm just under-caffeinated this morning ... I'm not
getting
the
notion of "average by Time".

Your UNION query sounds like one solution to being forced to use
text
files.
Another approach is to parse the data into a well-normalized
table
using
queries...

Regards

Jeff Boyce
<Office/Access MVP>


Hello Jeff

Thanks for your response.

Firstly, I've created a table that has all the "details":

Cell Treatment Start time
1 Pair 29/11/2005 11:54:59 AM
3 Manifold 29/11/2005 11:55:24 AM
6 Manifold 29/11/2005 11:55:47 AM
7 Pair 29/11/2005 11:56:13 AM
9 Pair 29/11/2005 11:56:35 AM

I agree with you that having one table per cell isn't great -
but
that's
how
I have to do it. I have to link to the txt files as outputted
from
the
instrument. I can not change the format of these files and
linking
to
them
allows me to keep them up to date (the instrument writes to the
txt
file
every half hour.

I have a union query which collates all the individual txt
files:
Hr Cell mL gas
0.5 1 0.2
0.5 3 0
0.5 4 2.1
0.5 6 0
0.5 7 0

I can then create the relationship between cell and treatment.

I can average the mL Gas when I use hour; but would like to be
able
to
do
it
with Time instead (and these become out of sync). Also,
occassionally
the
instrument hiccups and hours also become slightly out of sync
too.

I need to be able to average mL in time intervals, ideally half
hour
intervals but I can be flexiable on this.

Cheers

Susan

:

Susan

How do you know that cells 3, 15 & 24 are one "treatment"?
How
does
Access
know that? Unless you (or Access) has a way to connect those
together
as
one treatment, I don't see how you can average their
durations.

I didn't see your previous post, but wonder about a data
design
that
(apparently) uses one table per ?cell. If this is true, this
is
not
particularly normalized, and will make your analysis much
harder.

--
Regards

Jeff Boyce
<Office/Access MVP>

Hello,

I posted (creating relationships with time series)
yesterday
within
"access
Database Design".

Basically I have 24 tables that are linked to txt files
(the
output of
my
instrument). Yesterdays question concerned how to create
the
relationships
with these tables; and solved my problem.

I now want to...

My instrument measures mL of O2 delievered to 24 different
cells.
I
usually
run experiments where I duplicate "treatments" in 3 or more
cells.

My data base contains the linked txt files, a "details"
table
which
contains
the columns "cell, treatment, start time"; and the query
created
yesterday
which collates all of "hour" (time elapsed) and "mL gas"
and
cell
into
one
query.

I then created a query to calculate "date running"
[timeRuning]:[StartTime]+([hour]/24]).

I can scatterplot each cell RunningTime vrs mLgas.

What I would like to do is produce another scatter plot but
with
averages
of
mLgas, based on "treatment". For example cells 3, 15 and 24
are
one
"treatment". I can create averages of the mLgas.

My problem is that all the cells arn't started at the same
time,
I
would
like to be able to average them on a half hour basis. My
current
crosstab
query returns this

Time Manifold Old Style Pair
26/11/2005 11:00:00 PM 1.2 1.05 3.25
26/11/2005 11:00:36 PM 1.1
26/11/2005 11:01:12 PM 2.9
26/11/2005 11:01:48 PM 0
26/11/2005 11:02:24 PM 2.8 1.15

As can be seen in the "pair" column (a treatment) as each
cell
was
started
slightly different times, the result of the average doesn't
make
sense;
all
values should have been included in one row.

Can anyone help please???

Susan