Infinite Error Bars and how to correct them

[abcName-2]

I am having the issue or always receiving this message when plotting mean
chloride concentration against distance.

'One or more error bar calculations yielded infinite results. These error
bars have been removed from the chart.'

How can I fix this so I have the correct error bars?

<http://spssx-discussion.1045642.n5.nabble.com/file/t341671/stats.jpg>
<http://spssx-discussion.1045642.n5.nabble.com/file/t341671/stats2.jpg>
<http://spssx-discussion.1045642.n5.nabble.com/file/t341671/stats_3.jpg>




--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Jon Peck]

This message usually means that there is insufficient data to calculate that error bar, e. g. too few data points there.  In that case there is no “correct” error bar.

On Tue, Oct 8, 2019 at 7:11 AM abcName <[hidden email]> wrote:

I am having the issue or always receiving this message when plotting mean
chloride concentration against distance.

'One or more error bar calculations yielded infinite results. These error
bars have been removed from the chart.'

How can I fix this so I have the correct error bars?

<http://spssx-discussion.1045642.n5.nabble.com/file/t341671/stats.jpg>
<http://spssx-discussion.1045642.n5.nabble.com/file/t341671/stats2.jpg>
<http://spssx-discussion.1045642.n5.nabble.com/file/t341671/stats_3.jpg>




--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

--
===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[Andy W]

That plot type is trying to aggregate the data you gift it, so it is not
taking your already calculated SD's to make the error bars. Calculate where
you want the lines to go (lower and upper limits for your error bars), and
then put them into the chart using the High-Low option in the chart builder.

I have some examples on my site here,
https://andrewpwheeler.wordpress.com/2016/03/08/on-overlapping-error-bars-in-charts/
(and the code at
https://dl.dropboxusercontent.com/s/a4jz8cggdhhz8qb/Overlap_Graphs.sps?dl=0).
The very last GGRAPH command will be the closest to your current data, and
shows how you can use inline TRANS commands to calculate the lower and upper
ranges for your error bars.



-----
Andy W
[hidden email]
http://andrewpwheeler.wordpress.com/
--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD