scalebars

Matej_Tyc · November 18, 2012, 2:16pm

Dear developers,
I use Matplotlib to process and display images acquired by microscopes.
It is quite common to indicate dimensions by displaying scale bar in the
image rather than using axes with labels. Although axes enable you to
refer to specific location in the image, they take up space around the
image, so if you only need to show the scale, scale bar is better.

What is needed:
- The scale bar of given dimension (data units), possibly with bars at
its ends.
- Text (presumably centered under the bar), text size as well as
vertical offset in physical units (= units reflecting the actual image
size, like the font size)
- Semi-transparent rectangle, so the scale and label are more readable
- Dark/bright theme might be a good idea.
I have made an svg file in Inkscape, so you can see what I mean.

First of all, I tried to implement the stuff myself, but later I have
found out that there is something on github. I have forked it, made some
minor modifications, and I think that it is "almost done".
https://gist.github.com/4100881 (the add_scalebar function there is
broken ATM)
I also attach the test code for your convenience. You need to run it
with scalebars.py in the same directory.
You are supposed to see a tiny bright scalebar at the bottom right
corner.

There are some outstanding issues, though:

- I have a feeling that bars at the end of the scale bar should be
related to the font size, as well as the actual width of the scale bar.
How to achieve this?
- How to make the semi-transparent background for the bar and label in
a smart way?

Could you help me with those? I would like this to appear in matplotlib
since it is IMO a useful feature, what needs to be done?
Regards,
Matěj Týč

sbars-test.py (296 Bytes)

Joe_Kington · November 18, 2012, 5:32pm

Have you had a look at “AnchoredSizeBar” from mpl_toolkits.axes_grid1.anchored_artists?

http://matplotlib.org/1.1.1/mpl_toolkits/axes_grid/users/overview.html#anchoredartists

It provides essentially all of the features you mention. I’d agree it could use a few enhancements, but it’s a good start on this.

As a quick example of using it as you describe:

import matplotlib.pyplot as plt
import numpy as np
from mpl_toolkits.axes_grid1.anchored_artists import AnchoredSizeBar

fig, ax = plt.subplots()
ax.imshow(np.random.random((10,10)))

bar = AnchoredSizeBar(ax.transData, 2, ‘2 Units’,
pad=0.5, loc=8, sep=5, borderpad=0.5, frameon=True)
bar.patch.set(alpha=0.5, boxstyle=‘round’)
ax.add_artist(bar)

plt.show()

Note that this is very similar to your example. The main things it’s missing are ends on the scalebar.

I’d certainly agree that it could use some enhancements (e.g. different styles of scalebars and better documentation), but perhaps it’s best to start with AnchoredSizeBar instead of recreating it from scratch?

Just my thoughts, anyway.
-Joe

···

On Sun, Nov 18, 2012 at 8:16 AM, Matěj Týč <matej.tyc@…149…> wrote:

Dear developers,

I use Matplotlib to process and display images acquired by microscopes.

It is quite common to indicate dimensions by displaying scale bar in the

image rather than using axes with labels. Although axes enable you to

refer to specific location in the image, they take up space around the

image, so if you only need to show the scale, scale bar is better.

What is needed:

The scale bar of given dimension (data units), possibly with bars at

its ends.

Text (presumably centered under the bar), text size as well as

vertical offset in physical units (= units reflecting the actual image

size, like the font size)

Semi-transparent rectangle, so the scale and label are more readable

Dark/bright theme might be a good idea.

I have made an svg file in Inkscape, so you can see what I mean.

First of all, I tried to implement the stuff myself, but later I have

found out that there is something on github. I have forked it, made some

minor modifications, and I think that it is “almost done”.

https://gist.github.com/4100881 (the add_scalebar function there is

broken ATM)

I also attach the test code for your convenience. You need to run it

with scalebars.py in the same directory.

You are supposed to see a tiny bright scalebar at the bottom right

corner.

There are some outstanding issues, though:

I have a feeling that bars at the end of the scale bar should be

related to the font size, as well as the actual width of the scale bar.

How to achieve this?

How to make the semi-transparent background for the bar and label in

a smart way?

Could you help me with those? I would like this to appear in matplotlib

since it is IMO a useful feature, what needs to be done?

Regards,

Matěj Týč

Monitor your physical, virtual and cloud infrastructure from a single

web console. Get in-depth insight into apps, servers, databases, vmware,

SAP, cloud infrastructure, etc. Download 30-day Free Trial.

Pricing starts from $795 for 25 servers or applications!

http://p.sf.net/sfu/zoho_dev2dev_nov

Matplotlib-devel mailing list

Matplotlib-devel@lists.sourceforge.net

https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

Matej_Tyc · November 21, 2012, 12:12pm

Thank you for your fast reply,
you are right, AnchoredSizeBar has indeed almost all features I would
like. Or it definitely has the most important ones.
I have stumbled upon the page you refer, but I must have overlooked it.
An actual documentation of the function wouldn't hurt

Anyway, these features seem to be missing:
- Bar styles (bar width, bar endings wouldn't hurt either).
- Colors (bar, text, background).

I would like to look into it, but it is usually more efficient if more
experienced persons provide some pointers. So if you think that I should
know something before attempting to add the functionality, please let me
know.
For instance, do you think that the enhancements I have proposed make
sense and should be integrated into AnchoredSizeBar?

Matej

···

On 11/18/2012 06:32 PM, Joe Kington wrote:

Have you had a look at "AnchoredSizeBar" from mpl_toolkits.axes_grid1.anchored_artists?

http://matplotlib.org/1.1.1/mpl_toolkits/axes_grid/users/overview.html#anchoredartists

It provides essentially all of the features you mention. I'd agree it could use a few enhancements, but it's a good start on this.

As a quick example of using it as you describe:

import matplotlib.pyplot as plt
import numpy as np
from mpl_toolkits.axes_grid1.anchored_artists import AnchoredSizeBar

fig, ax = plt.subplots()
ax.imshow(np.random.random((10,10)))

bar = AnchoredSizeBar(ax.transData, 2, '2 Units',
pad=0.5, loc=8, sep=5, borderpad=0.5, frameon=True)
bar.patch.set(alpha=0.5, boxstyle='round')
ax.add_artist(bar)

plt.show()

Note that this is very similar to your example. The main things it's missing are ends on the scalebar.

I'd certainly agree that it could use some enhancements (e.g. different styles of scalebars and better documentation), but perhaps it's best to start with AnchoredSizeBar instead of recreating it from scratch?

Just my thoughts, anyway.
-Joe

On Sun, Nov 18, 2012 at 8:16 AM, Matěj Týč <matej.tyc@...149... > <mailto:matej.tyc@…149…>> wrote:

    Dear developers,
    I use Matplotlib to process and display images acquired by
    microscopes.
    It is quite common to indicate dimensions by displaying scale bar
    in the
    image rather than using axes with labels. Although axes enable you to
    refer to specific location in the image, they take up space around the
    image, so if you only need to show the scale, scale bar is better.

    What is needed:
     - The scale bar of given dimension (data units), possibly with
    bars at
    its ends.
     - Text (presumably centered under the bar), text size as well as
    vertical offset in physical units (= units reflecting the actual image
    size, like the font size)
     - Semi-transparent rectangle, so the scale and label are more
    readable
     - Dark/bright theme might be a good idea.
    I have made an svg file in Inkscape, so you can see what I mean.

    First of all, I tried to implement the stuff myself, but later I have
    found out that there is something on github. I have forked it,
    made some
    minor modifications, and I think that it is "almost done".
    matplotlib: add scale bars to axes - enhanced · GitHub (the add_scalebar function there is
    broken ATM)
    I also attach the test code for your convenience. You need to run it
    with scalebars.py in the same directory.
    You are supposed to see a tiny bright scalebar at the bottom right
    corner.

    There are some outstanding issues, though:

     - I have a feeling that bars at the end of the scale bar should be
    related to the font size, as well as the actual width of the scale
    bar.
    How to achieve this?
     - How to make the semi-transparent background for the bar and
    label in
    a smart way?

    Could you help me with those? I would like this to appear in
    matplotlib
    since it is IMO a useful feature, what needs to be done?
    Regards,
    Matěj Týč

    ------------------------------------------------------------------------------
    Monitor your physical, virtual and cloud infrastructure from a single
    web console. Get in-depth insight into apps, servers, databases,
    vmware,
    SAP, cloud infrastructure, etc. Download 30-day Free Trial.
    Pricing starts from $795 for 25 servers or applications!
    http://p.sf.net/sfu/zoho_dev2dev_nov
    _______________________________________________
    Matplotlib-devel mailing list
    Matplotlib-devel@lists.sourceforge.net
    <mailto:Matplotlib-devel@lists.sourceforge.net>
    matplotlib-devel List Signup and Options