[Met_help] [rt.rap.ucar.edu #95544] History for How to
Tatiana Burek via RT
met_help at ucar.edu
Tue Oct 13 12:11:49 MDT 2020
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hello,
I created a "Roly" plot using METViewer (please see the attachment). I
wonder if there is a way to mark the 1:1 line darker or more thick? Thank
you.
Binyu
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: How to
From: John Halley Gotway
Time: Mon Jun 08 13:01:19 2020
Binyu (and Tatiana),
Using the existing version of METviewer, you can draw straight lines
on
plots using the "Plot Script Commands" box on the "Formatting" tab.
This enables you to insert R commands to be executed after rendering
the
plot. The R function for drawing straight lines is called "abline()":
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
abline(0, 1) draws the 1-1 line.
abline(h=1) draws a horizontal line at 1.
abline(h=0) draws a horizontal line at 0.
abline(v=5) draws a vertical line at 5.
This generally works well and you can use additional options to make
the
line thicker, a different type, or a different color.
However, I tried doing this on a Reliability diagram and there's an
obvious
problem! See attached screenshot.
plot script command = abline(0, 1, col="black", lwd=3)
https://metviewer.nws.noaa.gov/
The line clearly extends beyond the plotting box. While this works
well for
the series template, it looks like we haven't tested it for the
reliability
template.
I'm going to assign this ticket to Tatina Burek, the METviewer
developer,
to see if there's any easy fix for this behavior.
Thanks,
John
On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
met_help at ucar.edu> wrote:
>
> Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> Transaction: Ticket created by binyu.wang at noaa.gov
> Queue: met_help
> Subject: How to
> Owner: Nobody
> Requestors: binyu.wang at noaa.gov
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
>
>
> Hello,
>
> I created a "Roly" plot using METViewer (please see the attachment).
I
> wonder if there is a way to mark the 1:1 line darker or more thick?
Thank
> you.
>
> Binyu
>
>
------------------------------------------------
Subject: How to
From: John Halley Gotway
Time: Mon Jun 08 13:04:20 2020
Binyu,
FYI, replacing "abline()" with a call to "segments()" produces better
results.
*segments(0, 0, 1, 1, col="black", lwd=3)*
John
On Mon, Jun 8, 2020 at 1:01 PM John Halley Gotway <johnhg at ucar.edu>
wrote:
> Binyu (and Tatiana),
>
> Using the existing version of METviewer, you can draw straight lines
on
> plots using the "Plot Script Commands" box on the "Formatting" tab.
> This enables you to insert R commands to be executed after rendering
the
> plot. The R function for drawing straight lines is called
"abline()":
>
>
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
>
> abline(0, 1) draws the 1-1 line.
> abline(h=1) draws a horizontal line at 1.
> abline(h=0) draws a horizontal line at 0.
> abline(v=5) draws a vertical line at 5.
>
> This generally works well and you can use additional options to make
the
> line thicker, a different type, or a different color.
> However, I tried doing this on a Reliability diagram and there's an
> obvious problem! See attached screenshot.
>
> plot script command = abline(0, 1, col="black", lwd=3)
>
> https://metviewer.nws.noaa.gov/
>
> The line clearly extends beyond the plotting box. While this works
well
> for the series template, it looks like we haven't tested it for the
> reliability template.
>
> I'm going to assign this ticket to Tatina Burek, the METviewer
developer,
> to see if there's any easy fix for this behavior.
>
> Thanks,
> John
>
> On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> met_help at ucar.edu> wrote:
>
>>
>> Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
>> Transaction: Ticket created by binyu.wang at noaa.gov
>> Queue: met_help
>> Subject: How to
>> Owner: Nobody
>> Requestors: binyu.wang at noaa.gov
>> Status: new
>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
>>
>>
>> Hello,
>>
>> I created a "Roly" plot using METViewer (please see the
attachment). I
>> wonder if there is a way to mark the 1:1 line darker or more thick?
Thank
>> you.
>>
>> Binyu
>>
>>
------------------------------------------------
Subject: How to
From: Tatiana Burek
Time: Mon Jun 08 13:08:22 2020
You can also use abline with this parameters:
abline(a=0, b=1, lty=1, col="black", cex=1, lwd=black, xpd=FALSE)
Tatiana
On Mon Jun 08 13:04:20 2020, johnhg wrote:
> Binyu,
>
> FYI, replacing "abline()" with a call to "segments()" produces
better
> results.
>
> *segments(0, 0, 1, 1, col="black", lwd=3)*
>
> John
>
> On Mon, Jun 8, 2020 at 1:01 PM John Halley Gotway <johnhg at ucar.edu>
> wrote:
>
> > Binyu (and Tatiana),
> >
> > Using the existing version of METviewer, you can draw straight
lines
> > on
> > plots using the "Plot Script Commands" box on the "Formatting"
tab.
> > This enables you to insert R commands to be executed after
rendering
> > the
> > plot. The R function for drawing straight lines is called
"abline()":
> >
> >
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
> >
> > abline(0, 1) draws the 1-1 line.
> > abline(h=1) draws a horizontal line at 1.
> > abline(h=0) draws a horizontal line at 0.
> > abline(v=5) draws a vertical line at 5.
> >
> > This generally works well and you can use additional options to
make
> > the
> > line thicker, a different type, or a different color.
> > However, I tried doing this on a Reliability diagram and there's
an
> > obvious problem! See attached screenshot.
> >
> > plot script command = abline(0, 1, col="black", lwd=3)
> >
> > https://metviewer.nws.noaa.gov/
> >
> > The line clearly extends beyond the plotting box. While this works
> > well
> > for the series template, it looks like we haven't tested it for
the
> > reliability template.
> >
> > I'm going to assign this ticket to Tatina Burek, the METviewer
> > developer,
> > to see if there's any easy fix for this behavior.
> >
> > Thanks,
> > John
> >
> > On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> > met_help at ucar.edu> wrote:
> >
> >>
> >> Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> >> Transaction: Ticket created by binyu.wang at noaa.gov
> >> Queue: met_help
> >> Subject: How to
> >> Owner: Nobody
> >> Requestors: binyu.wang at noaa.gov
> >> Status: new
> >> Ticket <URL:
> >> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
> >>
> >>
> >> Hello,
> >>
> >> I created a "Roly" plot using METViewer (please see the
attachment).
> >> I
> >> wonder if there is a way to mark the 1:1 line darker or more
thick?
> >> Thank
> >> you.
> >>
> >> Binyu
> >>
> >>
------------------------------------------------
Subject: How to
From: binyu.wang at noaa.gov
Time: Mon Jun 08 13:23:42 2020
Great, thank you, Tatiana and John.
On Mon, Jun 8, 2020 at 3:08 PM Tatiana Burek via RT
<met_help at ucar.edu>
wrote:
> You can also use abline with this parameters:
> abline(a=0, b=1, lty=1, col="black", cex=1, lwd=black, xpd=FALSE)
>
> Tatiana
> On Mon Jun 08 13:04:20 2020, johnhg wrote:
> > Binyu,
> >
> > FYI, replacing "abline()" with a call to "segments()" produces
better
> > results.
> >
> > *segments(0, 0, 1, 1, col="black", lwd=3)*
> >
> > John
> >
> > On Mon, Jun 8, 2020 at 1:01 PM John Halley Gotway
<johnhg at ucar.edu>
> > wrote:
> >
> > > Binyu (and Tatiana),
> > >
> > > Using the existing version of METviewer, you can draw straight
lines
> > > on
> > > plots using the "Plot Script Commands" box on the "Formatting"
tab.
> > > This enables you to insert R commands to be executed after
rendering
> > > the
> > > plot. The R function for drawing straight lines is called
"abline()":
> > >
> > >
>
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
> > >
> > > abline(0, 1) draws the 1-1 line.
> > > abline(h=1) draws a horizontal line at 1.
> > > abline(h=0) draws a horizontal line at 0.
> > > abline(v=5) draws a vertical line at 5.
> > >
> > > This generally works well and you can use additional options to
make
> > > the
> > > line thicker, a different type, or a different color.
> > > However, I tried doing this on a Reliability diagram and there's
an
> > > obvious problem! See attached screenshot.
> > >
> > > plot script command = abline(0, 1, col="black", lwd=3)
> > >
> > > https://metviewer.nws.noaa.gov/
> > >
> > > The line clearly extends beyond the plotting box. While this
works
> > > well
> > > for the series template, it looks like we haven't tested it for
the
> > > reliability template.
> > >
> > > I'm going to assign this ticket to Tatina Burek, the METviewer
> > > developer,
> > > to see if there's any easy fix for this behavior.
> > >
> > > Thanks,
> > > John
> > >
> > > On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > >>
> > >> Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> > >> Transaction: Ticket created by binyu.wang at noaa.gov
> > >> Queue: met_help
> > >> Subject: How to
> > >> Owner: Nobody
> > >> Requestors: binyu.wang at noaa.gov
> > >> Status: new
> > >> Ticket <URL:
> > >> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
> > >>
> > >>
> > >> Hello,
> > >>
> > >> I created a "Roly" plot using METViewer (please see the
attachment).
> > >> I
> > >> wonder if there is a way to mark the 1:1 line darker or more
thick?
> > >> Thank
> > >> you.
> > >>
> > >> Binyu
> > >>
> > >>
>
>
>
>
------------------------------------------------
Subject: How to
From: binyu.wang at noaa.gov
Time: Mon Jun 08 13:48:06 2020
Hello John,
There is another question for the "HIST" plot:
There is one thing I want to double check: all the statistics data
from MET
were obtained using the paired data, is that correct? Eg: if obs. have
valid data over 1500 grids, while forecast only has valid data over
1000
grids (miss 500 grids), so there are only 1000 paired grids. Then all
the
stat data (like RMSE etc) and plots (lile Hist) are calculated based
on the
1000 paired grids ONLY, is that correct?
There are 2 plots in the attachment which are made using the same
dataset.
In the second plot, the forecast (green color) seems to overestimate
obs
(orange color), but the Hist plot (the first attachment) looks
opposite.
It seems the forecast does not catch all obs., so maybe that is why
Hist
indicates an underestimate? But if HIst ONLY compares those matched
pairs,
then I don't know where the problem is. Any hint?
Thank you.
Binyu
On Mon, Jun 8, 2020 at 3:02 PM John Halley Gotway via RT
<met_help at ucar.edu>
wrote:
> Binyu (and Tatiana),
>
> Using the existing version of METviewer, you can draw straight lines
on
> plots using the "Plot Script Commands" box on the "Formatting" tab.
> This enables you to insert R commands to be executed after rendering
the
> plot. The R function for drawing straight lines is called
"abline()":
>
>
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
>
> abline(0, 1) draws the 1-1 line.
> abline(h=1) draws a horizontal line at 1.
> abline(h=0) draws a horizontal line at 0.
> abline(v=5) draws a vertical line at 5.
>
> This generally works well and you can use additional options to make
the
> line thicker, a different type, or a different color.
> However, I tried doing this on a Reliability diagram and there's an
obvious
> problem! See attached screenshot.
>
> plot script command = abline(0, 1, col="black", lwd=3)
>
> https://metviewer.nws.noaa.gov/
>
> The line clearly extends beyond the plotting box. While this works
well for
> the series template, it looks like we haven't tested it for the
reliability
> template.
>
> I'm going to assign this ticket to Tatina Burek, the METviewer
developer,
> to see if there's any easy fix for this behavior.
>
> Thanks,
> John
>
> On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> met_help at ucar.edu> wrote:
>
> >
> > Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> > Transaction: Ticket created by binyu.wang at noaa.gov
> > Queue: met_help
> > Subject: How to
> > Owner: Nobody
> > Requestors: binyu.wang at noaa.gov
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
> >
> >
> > Hello,
> >
> > I created a "Roly" plot using METViewer (please see the
attachment). I
> > wonder if there is a way to mark the 1:1 line darker or more
thick? Thank
> > you.
> >
> > Binyu
> >
> >
>
>
------------------------------------------------
Subject: How to
From: John Halley Gotway
Time: Mon Jun 08 15:02:36 2020
Binyu,
I would recommend looking closely at the ascii output files produced
by
Ensemble-Stat. In particular, the RHIST line type contains the counts
for
the ranked histogram data you're displaying. And the TOTAL column in
that
line type indicates the number of matched pairs that are included.
That
will tell you if you have 1500 matched pairs, 1000 matched pairs, or
some
other number.
If you'd like even more detail about which stations are included, you
can
re-configure and re-run Ensemble-Stat turning on the output_flag
option for
the ORANK line type. For each of the matched pairs, an ORANK line is
written to the output indicating the observation value, as well as all
the
values of all the ensemble members at that location. It's a very
useful way
to look at more detail in the data.
As for interpreting and guessing at the source of differences between
the
two plots you sent, I really don't know.
Thanks,
John
On Mon, Jun 8, 2020 at 1:48 PM binyu.wang at noaa.gov via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
>
> Hello John,
>
> There is another question for the "HIST" plot:
> There is one thing I want to double check: all the statistics data
from MET
> were obtained using the paired data, is that correct? Eg: if obs.
have
> valid data over 1500 grids, while forecast only has valid data over
1000
> grids (miss 500 grids), so there are only 1000 paired grids. Then
all the
> stat data (like RMSE etc) and plots (lile Hist) are calculated based
on the
> 1000 paired grids ONLY, is that correct?
>
> There are 2 plots in the attachment which are made using the same
dataset.
> In the second plot, the forecast (green color) seems to overestimate
obs
> (orange color), but the Hist plot (the first attachment) looks
opposite.
> It seems the forecast does not catch all obs., so maybe that is why
Hist
> indicates an underestimate? But if HIst ONLY compares those matched
pairs,
> then I don't know where the problem is. Any hint?
>
>
> Thank you.
> Binyu
>
>
>
> On Mon, Jun 8, 2020 at 3:02 PM John Halley Gotway via RT <
> met_help at ucar.edu>
> wrote:
>
> > Binyu (and Tatiana),
> >
> > Using the existing version of METviewer, you can draw straight
lines on
> > plots using the "Plot Script Commands" box on the "Formatting"
tab.
> > This enables you to insert R commands to be executed after
rendering the
> > plot. The R function for drawing straight lines is called
"abline()":
> >
> >
>
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
> >
> > abline(0, 1) draws the 1-1 line.
> > abline(h=1) draws a horizontal line at 1.
> > abline(h=0) draws a horizontal line at 0.
> > abline(v=5) draws a vertical line at 5.
> >
> > This generally works well and you can use additional options to
make the
> > line thicker, a different type, or a different color.
> > However, I tried doing this on a Reliability diagram and there's
an
> obvious
> > problem! See attached screenshot.
> >
> > plot script command = abline(0, 1, col="black", lwd=3)
> >
> > https://metviewer.nws.noaa.gov/
> >
> > The line clearly extends beyond the plotting box. While this works
well
> for
> > the series template, it looks like we haven't tested it for the
> reliability
> > template.
> >
> > I'm going to assign this ticket to Tatina Burek, the METviewer
developer,
> > to see if there's any easy fix for this behavior.
> >
> > Thanks,
> > John
> >
> > On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> > > Transaction: Ticket created by binyu.wang at noaa.gov
> > > Queue: met_help
> > > Subject: How to
> > > Owner: Nobody
> > > Requestors: binyu.wang at noaa.gov
> > > Status: new
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544
> >
> > >
> > >
> > > Hello,
> > >
> > > I created a "Roly" plot using METViewer (please see the
attachment). I
> > > wonder if there is a way to mark the 1:1 line darker or more
thick?
> Thank
> > > you.
> > >
> > > Binyu
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: How to
From: binyu.wang at noaa.gov
Time: Mon Jun 08 15:56:31 2020
John,
I didn't make the question clear in my previous email. I know how many
pairs based on the stat output. However I want to double check one
fact:
all the METVIEWer plots and stat (like MSE etc) are created ONLY based
on
the pairs, all the sites/events that were NOT catched by forecast
were
just ignored, is that correct? Thank you.
Binyu
On Mon, Jun 8, 2020 at 5:02 PM John Halley Gotway via RT
<met_help at ucar.edu>
wrote:
> Binyu,
>
> I would recommend looking closely at the ascii output files produced
by
> Ensemble-Stat. In particular, the RHIST line type contains the
counts for
> the ranked histogram data you're displaying. And the TOTAL column in
that
> line type indicates the number of matched pairs that are included.
That
> will tell you if you have 1500 matched pairs, 1000 matched pairs, or
some
> other number.
>
> If you'd like even more detail about which stations are included,
you can
> re-configure and re-run Ensemble-Stat turning on the output_flag
option for
> the ORANK line type. For each of the matched pairs, an ORANK line is
> written to the output indicating the observation value, as well as
all the
> values of all the ensemble members at that location. It's a very
useful way
> to look at more detail in the data.
>
> As for interpreting and guessing at the source of differences
between the
> two plots you sent, I really don't know.
>
> Thanks,
> John
>
> On Mon, Jun 8, 2020 at 1:48 PM binyu.wang at noaa.gov via RT <
> met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
> >
> > Hello John,
> >
> > There is another question for the "HIST" plot:
> > There is one thing I want to double check: all the statistics data
from
> MET
> > were obtained using the paired data, is that correct? Eg: if obs.
have
> > valid data over 1500 grids, while forecast only has valid data
over 1000
> > grids (miss 500 grids), so there are only 1000 paired grids. Then
all the
> > stat data (like RMSE etc) and plots (lile Hist) are calculated
based on
> the
> > 1000 paired grids ONLY, is that correct?
> >
> > There are 2 plots in the attachment which are made using the same
> dataset.
> > In the second plot, the forecast (green color) seems to
overestimate obs
> > (orange color), but the Hist plot (the first attachment) looks
opposite.
> > It seems the forecast does not catch all obs., so maybe that is
why Hist
> > indicates an underestimate? But if HIst ONLY compares those
matched
> pairs,
> > then I don't know where the problem is. Any hint?
> >
> >
> > Thank you.
> > Binyu
> >
> >
> >
> > On Mon, Jun 8, 2020 at 3:02 PM John Halley Gotway via RT <
> > met_help at ucar.edu>
> > wrote:
> >
> > > Binyu (and Tatiana),
> > >
> > > Using the existing version of METviewer, you can draw straight
lines on
> > > plots using the "Plot Script Commands" box on the "Formatting"
tab.
> > > This enables you to insert R commands to be executed after
rendering
> the
> > > plot. The R function for drawing straight lines is called
"abline()":
> > >
> > >
> >
>
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
> > >
> > > abline(0, 1) draws the 1-1 line.
> > > abline(h=1) draws a horizontal line at 1.
> > > abline(h=0) draws a horizontal line at 0.
> > > abline(v=5) draws a vertical line at 5.
> > >
> > > This generally works well and you can use additional options to
make
> the
> > > line thicker, a different type, or a different color.
> > > However, I tried doing this on a Reliability diagram and there's
an
> > obvious
> > > problem! See attached screenshot.
> > >
> > > plot script command = abline(0, 1, col="black", lwd=3)
> > >
> > > https://metviewer.nws.noaa.gov/
> > >
> > > The line clearly extends beyond the plotting box. While this
works well
> > for
> > > the series template, it looks like we haven't tested it for the
> > reliability
> > > template.
> > >
> > > I'm going to assign this ticket to Tatina Burek, the METviewer
> developer,
> > > to see if there's any easy fix for this behavior.
> > >
> > > Thanks,
> > > John
> > >
> > > On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> > > > Transaction: Ticket created by binyu.wang at noaa.gov
> > > > Queue: met_help
> > > > Subject: How to
> > > > Owner: Nobody
> > > > Requestors: binyu.wang at noaa.gov
> > > > Status: new
> > > > Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544
> > >
> > > >
> > > >
> > > > Hello,
> > > >
> > > > I created a "Roly" plot using METViewer (please see the
attachment).
> I
> > > > wonder if there is a way to mark the 1:1 line darker or more
thick?
> > Thank
> > > > you.
> > > >
> > > > Binyu
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: How to
From: John Halley Gotway
Time: Mon Jun 08 16:14:39 2020
Binyu,
Yes, that's correct. METviewer simply reads the RHIST output from
Ensemble-Stat and loads that into the database. And that's the data
that
shows up in the plot you're making.
And the MET tools only create matched pairs when both forecast and
observation data are present and contain valid data values. Any points
where the forecast or observation is either missing or contains bad
data
values are simple ignored.
John
On Mon, Jun 8, 2020 at 3:56 PM binyu.wang at noaa.gov via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
>
> John,
>
> I didn't make the question clear in my previous email. I know how
many
> pairs based on the stat output. However I want to double check one
fact:
> all the METVIEWer plots and stat (like MSE etc) are created ONLY
based on
> the pairs, all the sites/events that were NOT catched by forecast
were
> just ignored, is that correct? Thank you.
>
> Binyu
>
> On Mon, Jun 8, 2020 at 5:02 PM John Halley Gotway via RT <
> met_help at ucar.edu>
> wrote:
>
> > Binyu,
> >
> > I would recommend looking closely at the ascii output files
produced by
> > Ensemble-Stat. In particular, the RHIST line type contains the
counts for
> > the ranked histogram data you're displaying. And the TOTAL column
in that
> > line type indicates the number of matched pairs that are included.
That
> > will tell you if you have 1500 matched pairs, 1000 matched pairs,
or some
> > other number.
> >
> > If you'd like even more detail about which stations are included,
you can
> > re-configure and re-run Ensemble-Stat turning on the output_flag
option
> for
> > the ORANK line type. For each of the matched pairs, an ORANK line
is
> > written to the output indicating the observation value, as well as
all
> the
> > values of all the ensemble members at that location. It's a very
useful
> way
> > to look at more detail in the data.
> >
> > As for interpreting and guessing at the source of differences
between the
> > two plots you sent, I really don't know.
> >
> > Thanks,
> > John
> >
> > On Mon, Jun 8, 2020 at 1:48 PM binyu.wang at noaa.gov via RT <
> > met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
> > >
> > > Hello John,
> > >
> > > There is another question for the "HIST" plot:
> > > There is one thing I want to double check: all the statistics
data from
> > MET
> > > were obtained using the paired data, is that correct? Eg: if
obs. have
> > > valid data over 1500 grids, while forecast only has valid data
over
> 1000
> > > grids (miss 500 grids), so there are only 1000 paired grids.
Then all
> the
> > > stat data (like RMSE etc) and plots (lile Hist) are calculated
based on
> > the
> > > 1000 paired grids ONLY, is that correct?
> > >
> > > There are 2 plots in the attachment which are made using the
same
> > dataset.
> > > In the second plot, the forecast (green color) seems to
overestimate
> obs
> > > (orange color), but the Hist plot (the first attachment) looks
> opposite.
> > > It seems the forecast does not catch all obs., so maybe that is
why
> Hist
> > > indicates an underestimate? But if HIst ONLY compares those
matched
> > pairs,
> > > then I don't know where the problem is. Any hint?
> > >
> > >
> > > Thank you.
> > > Binyu
> > >
> > >
> > >
> > > On Mon, Jun 8, 2020 at 3:02 PM John Halley Gotway via RT <
> > > met_help at ucar.edu>
> > > wrote:
> > >
> > > > Binyu (and Tatiana),
> > > >
> > > > Using the existing version of METviewer, you can draw straight
lines
> on
> > > > plots using the "Plot Script Commands" box on the "Formatting"
tab.
> > > > This enables you to insert R commands to be executed after
rendering
> > the
> > > > plot. The R function for drawing straight lines is called
"abline()":
> > > >
> > > >
> > >
> >
>
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
> > > >
> > > > abline(0, 1) draws the 1-1 line.
> > > > abline(h=1) draws a horizontal line at 1.
> > > > abline(h=0) draws a horizontal line at 0.
> > > > abline(v=5) draws a vertical line at 5.
> > > >
> > > > This generally works well and you can use additional options
to make
> > the
> > > > line thicker, a different type, or a different color.
> > > > However, I tried doing this on a Reliability diagram and
there's an
> > > obvious
> > > > problem! See attached screenshot.
> > > >
> > > > plot script command = abline(0, 1, col="black", lwd=3)
> > > >
> > > > https://metviewer.nws.noaa.gov/
> > > >
> > > > The line clearly extends beyond the plotting box. While this
works
> well
> > > for
> > > > the series template, it looks like we haven't tested it for
the
> > > reliability
> > > > template.
> > > >
> > > > I'm going to assign this ticket to Tatina Burek, the METviewer
> > developer,
> > > > to see if there's any easy fix for this behavior.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> > > > > Transaction: Ticket created by binyu.wang at noaa.gov
> > > > > Queue: met_help
> > > > > Subject: How to
> > > > > Owner: Nobody
> > > > > Requestors: binyu.wang at noaa.gov
> > > > > Status: new
> > > > > Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544
> > > >
> > > > >
> > > > >
> > > > > Hello,
> > > > >
> > > > > I created a "Roly" plot using METViewer (please see the
> attachment).
> > I
> > > > > wonder if there is a way to mark the 1:1 line darker or more
thick?
> > > Thank
> > > > > you.
> > > > >
> > > > > Binyu
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: How to
From: binyu.wang at noaa.gov
Time: Tue Jun 09 14:02:35 2020
Hello Tatiana,
I have another question for you about making ROC plot using METviewer:
how
to change the range of x-axis? I tried to change the limit of x under
"x1
bound", but it doesn't work.
I tried to change the range of x-axis from (0-1) to (-0.1-1.1) as the
plot
"Figure C.2: Example of ROC Curve" on page 452" on the MET document.
https://dtcenter.org/sites/default/files/community-code/met/docs/user-
guide/MET_Users_Guide_v9.0.1.pdf
Thank you.
BInyu
On Mon, Jun 8, 2020 at 3:08 PM Tatiana Burek via RT
<met_help at ucar.edu>
wrote:
> You can also use abline with this parameters:
> abline(a=0, b=1, lty=1, col="black", cex=1, lwd=black, xpd=FALSE)
>
> Tatiana
> On Mon Jun 08 13:04:20 2020, johnhg wrote:
> > Binyu,
> >
> > FYI, replacing "abline()" with a call to "segments()" produces
better
> > results.
> >
> > *segments(0, 0, 1, 1, col="black", lwd=3)*
> >
> > John
> >
> > On Mon, Jun 8, 2020 at 1:01 PM John Halley Gotway
<johnhg at ucar.edu>
> > wrote:
> >
> > > Binyu (and Tatiana),
> > >
> > > Using the existing version of METviewer, you can draw straight
lines
> > > on
> > > plots using the "Plot Script Commands" box on the "Formatting"
tab.
> > > This enables you to insert R commands to be executed after
rendering
> > > the
> > > plot. The R function for drawing straight lines is called
"abline()":
> > >
> > >
>
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
> > >
> > > abline(0, 1) draws the 1-1 line.
> > > abline(h=1) draws a horizontal line at 1.
> > > abline(h=0) draws a horizontal line at 0.
> > > abline(v=5) draws a vertical line at 5.
> > >
> > > This generally works well and you can use additional options to
make
> > > the
> > > line thicker, a different type, or a different color.
> > > However, I tried doing this on a Reliability diagram and there's
an
> > > obvious problem! See attached screenshot.
> > >
> > > plot script command = abline(0, 1, col="black", lwd=3)
> > >
> > > https://metviewer.nws.noaa.gov/
> > >
> > > The line clearly extends beyond the plotting box. While this
works
> > > well
> > > for the series template, it looks like we haven't tested it for
the
> > > reliability template.
> > >
> > > I'm going to assign this ticket to Tatina Burek, the METviewer
> > > developer,
> > > to see if there's any easy fix for this behavior.
> > >
> > > Thanks,
> > > John
> > >
> > > On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > >>
> > >> Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> > >> Transaction: Ticket created by binyu.wang at noaa.gov
> > >> Queue: met_help
> > >> Subject: How to
> > >> Owner: Nobody
> > >> Requestors: binyu.wang at noaa.gov
> > >> Status: new
> > >> Ticket <URL:
> > >> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
> > >>
> > >>
> > >> Hello,
> > >>
> > >> I created a "Roly" plot using METViewer (please see the
attachment).
> > >> I
> > >> wonder if there is a way to mark the 1:1 line darker or more
thick?
> > >> Thank
> > >> you.
> > >>
> > >> Binyu
> > >>
> > >>
>
>
>
>
------------------------------------------------
Subject: How to
From: Tatiana Burek
Time: Wed Jun 10 07:42:30 2020
BInyu
METviewer doesn't support this feature for now.
Tatiana
On Tue Jun 09 14:02:35 2020, binyu.wang at noaa.gov wrote:
> Hello Tatiana,
>
> I have another question for you about making ROC plot using
METviewer:
> how
> to change the range of x-axis? I tried to change the limit of x
under
> "x1
> bound", but it doesn't work.
>
> I tried to change the range of x-axis from (0-1) to (-0.1-1.1) as
the
> plot
> "Figure C.2: Example of ROC Curve" on page 452" on the MET
document.
>
> https://dtcenter.org/sites/default/files/community-
code/met/docs/user-
> guide/MET_Users_Guide_v9.0.1.pdf
>
>
> Thank you.
> BInyu
>
> On Mon, Jun 8, 2020 at 3:08 PM Tatiana Burek via RT
> <met_help at ucar.edu>
> wrote:
>
> > You can also use abline with this parameters:
> > abline(a=0, b=1, lty=1, col="black", cex=1, lwd=black,
xpd=FALSE)
> >
> > Tatiana
> > On Mon Jun 08 13:04:20 2020, johnhg wrote:
> > > Binyu,
> > >
> > > FYI, replacing "abline()" with a call to "segments()" produces
> > > better
> > > results.
> > >
> > > *segments(0, 0, 1, 1, col="black", lwd=3)*
> > >
> > > John
> > >
> > > On Mon, Jun 8, 2020 at 1:01 PM John Halley Gotway
<johnhg at ucar.edu>
> > > wrote:
> > >
> > > > Binyu (and Tatiana),
> > > >
> > > > Using the existing version of METviewer, you can draw straight
> > > > lines
> > > > on
> > > > plots using the "Plot Script Commands" box on the "Formatting"
> > > > tab.
> > > > This enables you to insert R commands to be executed after
> > > > rendering
> > > > the
> > > > plot. The R function for drawing straight lines is called
> > > > "abline()":
> > > >
> > > >
> >
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
> > > >
> > > > abline(0, 1) draws the 1-1 line.
> > > > abline(h=1) draws a horizontal line at 1.
> > > > abline(h=0) draws a horizontal line at 0.
> > > > abline(v=5) draws a vertical line at 5.
> > > >
> > > > This generally works well and you can use additional options
to
> > > > make
> > > > the
> > > > line thicker, a different type, or a different color.
> > > > However, I tried doing this on a Reliability diagram and
there's
> > > > an
> > > > obvious problem! See attached screenshot.
> > > >
> > > > plot script command = abline(0, 1, col="black", lwd=3)
> > > >
> > > > https://metviewer.nws.noaa.gov/
> > > >
> > > > The line clearly extends beyond the plotting box. While this
> > > > works
> > > > well
> > > > for the series template, it looks like we haven't tested it
for
> > > > the
> > > > reliability template.
> > > >
> > > > I'm going to assign this ticket to Tatina Burek, the METviewer
> > > > developer,
> > > > to see if there's any easy fix for this behavior.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > >>
> > > >> Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> > > >> Transaction: Ticket created by binyu.wang at noaa.gov
> > > >> Queue: met_help
> > > >> Subject: How to
> > > >> Owner: Nobody
> > > >> Requestors: binyu.wang at noaa.gov
> > > >> Status: new
> > > >> Ticket <URL:
> > > >> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
> > > >>
> > > >>
> > > >> Hello,
> > > >>
> > > >> I created a "Roly" plot using METViewer (please see the
> > > >> attachment).
> > > >> I
> > > >> wonder if there is a way to mark the 1:1 line darker or more
> > > >> thick?
> > > >> Thank
> > > >> you.
> > > >>
> > > >> Binyu
> > > >>
> > > >>
> >
> >
> >
> >
------------------------------------------------
Subject: How to
From: binyu.wang at noaa.gov
Time: Wed Jun 10 08:27:22 2020
Tatiana,
Thank you for your reply. Is the plot in the document created using
METPLUS?
Binyu
On Wed, Jun 10, 2020 at 9:42 AM Tatiana Burek via RT
<met_help at ucar.edu>
wrote:
> BInyu
>
> METviewer doesn't support this feature for now.
>
> Tatiana
>
> On Tue Jun 09 14:02:35 2020, binyu.wang at noaa.gov wrote:
> > Hello Tatiana,
> >
> > I have another question for you about making ROC plot using
METviewer:
> > how
> > to change the range of x-axis? I tried to change the limit of x
under
> > "x1
> > bound", but it doesn't work.
> >
> > I tried to change the range of x-axis from (0-1) to (-0.1-1.1) as
the
> > plot
> > "Figure C.2: Example of ROC Curve" on page 452" on the MET
document.
> >
> > https://dtcenter.org/sites/default/files/community-
code/met/docs/user-
> > guide/MET_Users_Guide_v9.0.1.pdf
> >
> >
> > Thank you.
> > BInyu
> >
> > On Mon, Jun 8, 2020 at 3:08 PM Tatiana Burek via RT
> > <met_help at ucar.edu>
> > wrote:
> >
> > > You can also use abline with this parameters:
> > > abline(a=0, b=1, lty=1, col="black", cex=1, lwd=black,
xpd=FALSE)
> > >
> > > Tatiana
> > > On Mon Jun 08 13:04:20 2020, johnhg wrote:
> > > > Binyu,
> > > >
> > > > FYI, replacing "abline()" with a call to "segments()" produces
> > > > better
> > > > results.
> > > >
> > > > *segments(0, 0, 1, 1, col="black", lwd=3)*
> > > >
> > > > John
> > > >
> > > > On Mon, Jun 8, 2020 at 1:01 PM John Halley Gotway
<johnhg at ucar.edu>
> > > > wrote:
> > > >
> > > > > Binyu (and Tatiana),
> > > > >
> > > > > Using the existing version of METviewer, you can draw
straight
> > > > > lines
> > > > > on
> > > > > plots using the "Plot Script Commands" box on the
"Formatting"
> > > > > tab.
> > > > > This enables you to insert R commands to be executed after
> > > > > rendering
> > > > > the
> > > > > plot. The R function for drawing straight lines is called
> > > > > "abline()":
> > > > >
> > > > >
> > >
>
https://www.rdocumentation.org/packages/graphics/versions/3.6.2/topics/abline
> > > > >
> > > > > abline(0, 1) draws the 1-1 line.
> > > > > abline(h=1) draws a horizontal line at 1.
> > > > > abline(h=0) draws a horizontal line at 0.
> > > > > abline(v=5) draws a vertical line at 5.
> > > > >
> > > > > This generally works well and you can use additional options
to
> > > > > make
> > > > > the
> > > > > line thicker, a different type, or a different color.
> > > > > However, I tried doing this on a Reliability diagram and
there's
> > > > > an
> > > > > obvious problem! See attached screenshot.
> > > > >
> > > > > plot script command = abline(0, 1, col="black", lwd=3)
> > > > >
> > > > > https://metviewer.nws.noaa.gov/
> > > > >
> > > > > The line clearly extends beyond the plotting box. While this
> > > > > works
> > > > > well
> > > > > for the series template, it looks like we haven't tested it
for
> > > > > the
> > > > > reliability template.
> > > > >
> > > > > I'm going to assign this ticket to Tatina Burek, the
METviewer
> > > > > developer,
> > > > > to see if there's any easy fix for this behavior.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Mon, Jun 8, 2020 at 12:48 PM binyu.wang at noaa.gov via RT <
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > >>
> > > > >> Mon Jun 08 12:48:19 2020: Request 95544 was acted upon.
> > > > >> Transaction: Ticket created by binyu.wang at noaa.gov
> > > > >> Queue: met_help
> > > > >> Subject: How to
> > > > >> Owner: Nobody
> > > > >> Requestors: binyu.wang at noaa.gov
> > > > >> Status: new
> > > > >> Ticket <URL:
> > > > >> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95544 >
> > > > >>
> > > > >>
> > > > >> Hello,
> > > > >>
> > > > >> I created a "Roly" plot using METViewer (please see the
> > > > >> attachment).
> > > > >> I
> > > > >> wonder if there is a way to mark the 1:1 line darker or
more
> > > > >> thick?
> > > > >> Thank
> > > > >> you.
> > > > >>
> > > > >> Binyu
> > > > >>
> > > > >>
> > >
> > >
> > >
> > >
>
>
>
>
------------------------------------------------
More information about the Met_help
mailing list