Scientific Numbers in Scatter Plot
Friday, 25 January 2013 · 4 messages
IG
IVES Greg25 Jan 2013 13:38 #
Hi
I find BaseX a very powerful tool.
One thing I noticed however is that it misunderstands numbers in scientific notation in the scatter plot display. Please see the attached plot.
The samples should form a rough diagonal line.
However, look at the scale on the left and you can see the exponent field is 'dis-continuous'. It looks like the chart formatter treats it as a non-scientific number or possibly even lexicographic order.
Best Regards
Greg
LK
Lukas Kircher25 Jan 2013 14:13 #
Hi Greg,
the scatter plot takes its information whether an attribute is numerical/categorical from
some DB statistics. The checkbox for the log scale in your provided screenshot is
deactivated - most likely because BaseX assumes that the TIME attribute is categorical.
This explains the lexicographic order and the different casing (e-01 / E-01).
A quick fix could be to convert the exponential to decimal representation (if you need the
plot at the moment). For now I'm filing an [issue] on GitHub.
Is it possible that you supply us with a part of the document that reproduces the issue?
Regards,
Lukas
[issue] https://github.com/BaseXdb/basex/issues/626
On Fri, Jan 25, 2013 at 1:38 PM, IVES Greg <greg.ives at uk.thalesgroup.com>wrote:
Quoted text · IVES Greg · 31 lines
> Hi**** > > I find BaseX a very powerful tool.**** > > One thing I noticed however is that it misunderstands numbers in > scientific notation in the scatter plot display.**** > > Please see the attached plot.**** > > The samples should form a rough diagonal line.**** > > However, look at the scale on the left and you can see the exponent field > is 'dis-continuous'.**** > > It looks like the chart formatter treats it as a non-scientific number or > possibly even lexicographic order.**** > > ** ** > > ** ** > > Best Regards**** > > Greg**** > > _______________________________________________ > BaseX-Talk mailing list > BaseX-Talk at mailman.uni-konstanz.de > https://mailman.uni-konstanz.de/mailman/listinfo/basex-talk > >
IG
IVES Greg25 Jan 2013 14:19 #
Hi Lukas,
Thanks for the quick reply.
Yes I did in XQuery a subtract of 0 and made a new XML DB to work around this. The document has some sensitivity - I can create a new DB on only the field that we are interested in here if you want? Bear with me and I'll send it later today if so.
Best Regards
Greg
Quoted text · Lukas Kircher · 29 lines
From: Lukas Kircher [mailto:lukaskircher1 at gmail.com] Sent: 25 January 2013 13:13 To: IVES Greg Cc: basex-talk at mailman.uni-konstanz.de Subject: Re: [basex-talk] Scientific Numbers in Scatter Plot Hi Greg, the scatter plot takes its information whether an attribute is numerical/categorical from some DB statistics. The checkbox for the log scale in your provided screenshot is deactivated - most likely because BaseX assumes that the TIME attribute is categorical. This explains the lexicographic order and the different casing (e-01 / E-01). A quick fix could be to convert the exponential to decimal representation (if you need the plot at the moment). For now I'm filing an [issue] on GitHub. Is it possible that you supply us with a part of the document that reproduces the issue? Regards, Lukas [issue] https://github.com/BaseXdb/basex/issues/626 On Fri, Jan 25, 2013 at 1:38 PM, IVES Greg <greg.ives at uk.thalesgroup.com<mailto:greg.ives at uk.thalesgroup.com>> wrote: Hi I find BaseX a very powerful tool. One thing I noticed however is that it misunderstands numbers in scientific notation in the scatter plot display. Please see the attached plot. The samples should form a rough diagonal line. However, look at the scale on the left and you can see the exponent field is 'dis-continuous'. It looks like the chart formatter treats it as a non-scientific number or possibly even lexicographic order. Best Regards Greg BaseX-Talk mailing list BaseX-Talk at mailman.uni-konstanz.de<mailto:BaseX-Talk at mailman.uni-konstanz.de> https://mailman.uni-konstanz.de/mailman/listinfo/basex-talk
IG
Lukas
Decided to do this now:
I think this does the job -attached.
TIME is the problem field in a plot - (y-axis in previous picture).
Best Regards
Greg
Quoted text · IVES Greg · 41 lines
From: IVES Greg Sent: 25 January 2013 13:19 To: 'Lukas Kircher' Cc: basex-talk at mailman.uni-konstanz.de Subject: RE: [basex-talk] Scientific Numbers in Scatter Plot Hi Lukas, Thanks for the quick reply. Yes I did in XQuery a subtract of 0 and made a new XML DB to work around this. The document has some sensitivity - I can create a new DB on only the field that we are interested in here if you want? Bear with me and I'll send it later today if so. Best Regards Greg From: Lukas Kircher [mailto:lukaskircher1 at gmail.com] Sent: 25 January 2013 13:13 To: IVES Greg Cc: basex-talk at mailman.uni-konstanz.de Subject: Re: [basex-talk] Scientific Numbers in Scatter Plot Hi Greg, the scatter plot takes its information whether an attribute is numerical/categorical from some DB statistics. The checkbox for the log scale in your provided screenshot is deactivated - most likely because BaseX assumes that the TIME attribute is categorical. This explains the lexicographic order and the different casing (e-01 / E-01). A quick fix could be to convert the exponential to decimal representation (if you need the plot at the moment). For now I'm filing an [issue] on GitHub. Is it possible that you supply us with a part of the document that reproduces the issue? Regards, Lukas [issue] https://github.com/BaseXdb/basex/issues/626 On Fri, Jan 25, 2013 at 1:38 PM, IVES Greg <greg.ives at uk.thalesgroup.com<mailto:greg.ives at uk.thalesgroup.com>> wrote: Hi I find BaseX a very powerful tool. One thing I noticed however is that it misunderstands numbers in scientific notation in the scatter plot display. Please see the attached plot. The samples should form a rough diagonal line. However, look at the scale on the left and you can see the exponent field is 'dis-continuous'. It looks like the chart formatter treats it as a non-scientific number or possibly even lexicographic order. Best Regards Greg BaseX-Talk mailing list BaseX-Talk at mailman.uni-konstanz.de<mailto:BaseX-Talk at mailman.uni-konstanz.de> https://mailman.uni-konstanz.de/mailman/listinfo/basex-talk
Written for BaseX 7.5. BaseX may have changed since: please check the current documentation.
⚡Generated with XQuery