Improve docs example in first_steps_8.rst

[cdeil]

This is a tiny improvement to a docs example, adding a show() call to have a full working example and be consistent with the other examples in the getting started guide.

Filing a [BUG] for such small docs improvements that aren't really bugs seems overkill. Hope it's OK as an exception.

[codecov]

Codecov Report

Merging #13161 (4546f65) into branch-3.2 (bb3a605) will decrease coverage by 0.01%.
The diff coverage is n/a.

@@              Coverage Diff               @@
##           branch-3.2   #13161      +/-   ##
==============================================
- Coverage       92.42%   92.41%   -0.01%     
==============================================
  Files             316      315       -1     
  Lines           20040    20038       -2     
==============================================
- Hits            18521    18518       -3     
- Misses           1519     1520       +1     

[bryevdv]

cc @tcmetzger IIRC omitting the show here was an intentional choice.

[tcmetzger]

I honestly don't remember why there is no show() in this example, I don't really see a reason why it shouldn't be added. This example looks quite boring if rendered, and the point of using a different data structure as input isn't really illustrated by displaying the plot, but none of that should stop us from having show() in there for consistency.

If we do add show(), how about we make the figure a little smaller (by defining height=250) and make the circles a little bigger (size=20), for example like this:

from bokeh.plotting import figure, show
from bokeh.models import ColumnDataSource
# create dict as basis for ColumnDataSource
data = {'x_values': [1, 2, 3, 4, 5],
        'y_values': [6, 7, 2, 3, 6]}
# create ColumnDataSource based on dict
source = ColumnDataSource(data=data)
# create a plot and renderer with ColumnDataSource data
p = figure(height=250)
p.circle(x='x_values', y='y_values', size=20, source=source)
show(p)

[cdeil]

@tcmetzger - I applied your suggestion in 4546f65

[cdeil]

the point of using a different data structure as input isn't really illustrated by displaying the plot

True. The key question isn't explained: "Why does ColumnDataSource and the other classes exist? Even if they exist internally, why should I as user care or use them?"

none of that should stop us from having show() in there for consistency

I think simply having full working examples that newbies can copy & paste is the biggest advantage. This actually happened to me: I copy & pasted the example and didn't get a plot. Took me a minute to figure out that I was simply missing the "show" call, but the example otherwise was working perfectly.

---

I do think it would be good to have some information on "why" ColumnDataSource and the filters exist, and "when" I have to or should use them over simple pandas DataFrames, and also what the performance implications are when bokeh makes copies. But that's a different possible aspect of the docs that could be improved. At least I don't know the answer and can't add it here (if you agree something could be added).

[bryevdv]

Thanks for the PR @cdeil !