Bokek Hover Tool Displaying Incorrectly | Python

Community Support
1

Hello!

I have the following dataframe:

Date Impressions Link_Clicks

2017-05-29 00-00-00 1,492,046 8,093

2017-06-05 00-00-00 845,012 4,864

2017-06-12 00-00-00 1,167,100 6,897

2017-06-19 00-00-00 895,781 4,472

2017-06-26 00-00-00 1,037,839 9,518

2017-07-03 00-00-00 1,139,060 9,668

2017-07-10 00-00-00 1,235,760 9,268

2017-07-17 00-00-00 1,200,906 7,989

2017-07-24 00-00-00 1,214,534 6,991

2017-07-31 00-00-00 1,434,225 7,311

2017-08-07 00-00-00 557,393 2,908

I am trying to visualise this using Bokeh which I’ve managed to do successfully, however when implementing the hover-tool, it returns the number divided specifically by 2. Is this a bug?

Libraries used:

from bokeh.io import output_notebook, show, push_notebook

from bokeh.io import output_file, show, curdoc

output_notebook()

from bokeh.plotting import figure, ColumnDataSource

from bokeh.models import HoverTool, DatetimeTickFormatter, DataRange1d, CustomJS, Plot, LinearAxis, Grid

from bokeh.charts import Bar, BoxPlot, Donut, HeatMap, Histogram, Line, Scatter, TimeSeries

My code can be found below:

source = ColumnDataSource(data=dict(

desc = bar_df.index,

y = bar_df[“Link_Clicks”],

))

p = Bar(bar_df, label=bar_df.index, values=“Link_Clicks”, width=1, source=source, agg=‘sum’,

line_color=“white”, plot_width=900, plot_height=400, bar_width=0.9, legend=None, toolbar_location=“right”)

p.add_tools(HoverTool(tooltips=[(“Link_Clicks”, “@y{1.11}”),

(“Date”, “@index”)]))

p.xaxis.major_label_orientation = 45

show(p)

NOTE: if you try and replicate this, dates for Bokeh can’t be in DateTime format; only strings.

To quickly convert: df[“Date”] = df[“Date”].apply(lambda x: str(x).replace(’:’,’-’))

When hovering over, you can see this is divided perfectly by 2 which is strange. I have tried multiplying y*2 under ColumnDataSource and within the Hover Tool, but still no luck.

Any ideas?

2

Hi,

Don't use bokeh.charts (including Bar) for anything. It was a failed experiment, and was deprcated and subsequently removed entirely last year. It is currently completely unmaintained and unsupported.

Instead, use the well-supported bokeh.plotting API, which has had very significant improvements to make all kinds o bar charts much simpler (the bokeh.plotting API is now more capable than bokeh.charts.Bar ever was):

  http://bokeh.pydata.org/en/latest/docs/user_guide/categorical.html

Thanks,

Bryan

···

On May 22, 2018, at 08:02, Adrian Ciconte <[email protected]> wrote:

Hello!

I have the following dataframe:

Date Impressions Link_Clicks
2017-05-29 00-00-00 1,492,046 8,093
2017-06-05 00-00-00 845,012 4,864
2017-06-12 00-00-00 1,167,100 6,897
2017-06-19 00-00-00 895,781 4,472
2017-06-26 00-00-00 1,037,839 9,518
2017-07-03 00-00-00 1,139,060 9,668
2017-07-10 00-00-00 1,235,760 9,268
2017-07-17 00-00-00 1,200,906 7,989
2017-07-24 00-00-00 1,214,534 6,991
2017-07-31 00-00-00 1,434,225 7,311
2017-08-07 00-00-00 557,393 2,908

I am trying to visualise this using Bokeh which I've managed to do successfully, however when implementing the hover-tool, it returns the number divided specifically by 2. Is this a bug?

Libraries used:

from bokeh.io import output_notebook, show, push_notebook
from bokeh.io import output_file, show, curdoc
output_notebook()
from bokeh.plotting import figure, ColumnDataSource
from bokeh.models import HoverTool, DatetimeTickFormatter, DataRange1d, CustomJS, Plot, LinearAxis, Grid
from bokeh.charts import Bar, BoxPlot, Donut, HeatMap, Histogram, Line, Scatter, TimeSeries

My code can be found below:

source = ColumnDataSource(data=dict(
    desc = bar_df.index,
    y = bar_df["Link_Clicks"],
))

p = Bar(bar_df, label=bar_df.index, values="Link_Clicks", width=1, source=source, agg='sum',
       line_color="white", plot_width=900, plot_height=400, bar_width=0.9, legend=None, toolbar_location="right")

p.add_tools(HoverTool(tooltips=[("Link_Clicks", "@y{1.11}"),
                                ("Date", "@index")]))

p.xaxis.major_label_orientation = 45

show(p)

NOTE: if you try and replicate this, dates for Bokeh can't be in DateTime format; only strings.
To quickly convert: df["Date"] = df["Date"].apply(lambda x: str(x).replace(':','-'))

When hovering over, you can see this is divided perfectly by 2 which is strange. I have tried multiplying y*2 under ColumnDataSource and within the Hover Tool, but still no luck.

Any ideas?

--
You received this message because you are subscribed to the Google Groups "Bokeh Discussion - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit https://groups.google.com/a/continuum.io/d/msgid/bokeh/3f680b76-46b0-467f-a5a0-4d172d96b9dd%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/d/optout\.

3

Hi,

You are sending a literal list or array as the "x" value:

  x=bokeh_bar["Audience"]

But then sending other values as columns from a CDS source:

  color='color', ..., source=source

If you use a "source" arg at all, then all values need to be in the CDS source, as the error states. The solution is to put the "x" values in your CDS as well.

Thanks,

Bryan

···

On May 22, 2018, at 18:57, Adrian Ciconte <[email protected]> wrote:

Thanks Bryan,

I have been reading over the documentation and applying the plotting API syntax (based on the documentation you sent) to my problem (which worked perfectly using version 0.12.02) however, when I upgraded to 0.12.16, I get the following error.

Data:

bokeh_bar = df.groupby(["Audience"], as_index=False)["Impressions"].sum()

Index Audience Impressions
0 Core 23725548
1 Homepage 14163811
2 LAL_10% 18277859
3 LAL_3% 14879857
4 Page_1 4406266

Code:

source = ColumnDataSource(data=dict(audience=bokeh_bar["Audience"],
                                    impressions=bokeh_bar["Impressions"],
                                    color=Spectral6))

p = figure(x_range=list(bokeh_bar["Audience"]), plot_height=250, title="Impressions by Audience",
           toolbar_location=None, tools="")

p.vbar(x=bokeh_bar["Audience"], top="Impressions", width=0.9, color='color',
       legend="Audience", source=source)

p.add_tools(HoverTool(tooltips=[("Audience", "@audience"), ("Impressions", "@impressions")]))

p.xgrid.grid_line_color = None
p.legend.orientation = "horizontal"
p.legend.location = "top_right"

show(p)

Error:

RuntimeError:
Supplying a user-defined data source AND iterable values to glyph methods is
not possibe. Either:

Pass all data directly as literals:

    p.circe(x=a_list, y=an_array, ...)

Or, put all data in a ColumnDataSource and pass column names:

    source = ColumnDataSource(data=dict(x=a_list, y=an_array))
    p.circe(x='x', y='y', source=source, ...)

Am I missing something glaringly obvious? I feel like I've addressed both these errors in my code.

Thanks in advance for your help.

Adrian

On Wednesday, May 23, 2018 at 2:02:32 AM UTC+10, Bryan Van de ven wrote:
Hi,

Don't use bokeh.charts (including Bar) for anything. It was a failed experiment, and was deprcated and subsequently removed entirely last year. It is currently completely unmaintained and unsupported.

Instead, use the well-supported bokeh.plotting API, which has had very significant improvements to make all kinds o bar charts much simpler (the bokeh.plotting API is now more capable than bokeh.charts.Bar ever was):

        http://bokeh.pydata.org/en/latest/docs/user_guide/categorical.html

Thanks,

Bryan

> On May 22, 2018, at 08:02, Adrian Ciconte <[email protected]> wrote:
>
> Hello!
>
> I have the following dataframe:
>
> Date Impressions Link_Clicks
> 2017-05-29 00-00-00 1,492,046 8,093
> 2017-06-05 00-00-00 845,012 4,864
> 2017-06-12 00-00-00 1,167,100 6,897
> 2017-06-19 00-00-00 895,781 4,472
> 2017-06-26 00-00-00 1,037,839 9,518
> 2017-07-03 00-00-00 1,139,060 9,668
> 2017-07-10 00-00-00 1,235,760 9,268
> 2017-07-17 00-00-00 1,200,906 7,989
> 2017-07-24 00-00-00 1,214,534 6,991
> 2017-07-31 00-00-00 1,434,225 7,311
> 2017-08-07 00-00-00 557,393 2,908
>
> I am trying to visualise this using Bokeh which I've managed to do successfully, however when implementing the hover-tool, it returns the number divided specifically by 2. Is this a bug?
>
> Libraries used:
>
> from bokeh.io import output_notebook, show, push_notebook
> from bokeh.io import output_file, show, curdoc
> output_notebook()
> from bokeh.plotting import figure, ColumnDataSource
> from bokeh.models import HoverTool, DatetimeTickFormatter, DataRange1d, CustomJS, Plot, LinearAxis, Grid
> from bokeh.charts import Bar, BoxPlot, Donut, HeatMap, Histogram, Line, Scatter, TimeSeries
>
> My code can be found below:
>
> source = ColumnDataSource(data=dict(
> desc = bar_df.index,
> y = bar_df["Link_Clicks"],
> ))
>
> p = Bar(bar_df, label=bar_df.index, values="Link_Clicks", width=1, source=source, agg='sum',
> line_color="white", plot_width=900, plot_height=400, bar_width=0.9, legend=None, toolbar_location="right")
>
> p.add_tools(HoverTool(tooltips=[("Link_Clicks", "@y{1.11}"),
> ("Date", "@index")]))
>
> p.xaxis.major_label_orientation = 45
>
> show(p)
>
> NOTE: if you try and replicate this, dates for Bokeh can't be in DateTime format; only strings.
> To quickly convert: df["Date"] = df["Date"].apply(lambda x: str(x).replace(':','-'))
>
> When hovering over, you can see this is divided perfectly by 2 which is strange. I have tried multiplying y*2 under ColumnDataSource and within the Hover Tool, but still no luck.
>
> Any ideas?
>
> --
> You received this message because you are subscribed to the Google Groups "Bokeh Discussion - Public" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to bokeh+un...@continuum.io.
> To post to this group, send email to bo...@continuum.io.
> To view this discussion on the web visit https://groups.google.com/a/continuum.io/d/msgid/bokeh/3f680b76-46b0-467f-a5a0-4d172d96b9dd%40continuum.io\.
> For more options, visit https://groups.google.com/a/continuum.io/d/optout\.

--
You received this message because you are subscribed to the Google Groups "Bokeh Discussion - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit https://groups.google.com/a/continuum.io/d/msgid/bokeh/5c070f25-180e-485e-8b08-3083b096b3f6%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/d/optout\.