question on texas_example and possibility of key

Community Support
1

I looked at the example:
http://bokeh.pydata.org/plot_gallery/texas_example.html

and found the us_states.py and have successfully adapted it to my use case.

My question is - in the example there are five colors specifically chosen:

colors = ["#F1EEF6", “#D4B9DA”, “#C994C7”, “#DF65B0”, “#DD1C77”, “#980043”]

and the rate is later then forced into one of 5 buckets to choose the color:

idx = min(int(rate*100), 5)

Is there a way to not force it to the five colors, but to choose a gradient, then the scores can just go along the color line from 100 to 0 and choose the appropriate color?

Also, wouldn’t this then allow for a key to be place in the plot - is that possible?

2

Looking through the source code I believe I am just seeing that keys are not yet implemented.

So I will ask another question. In the geo plot I linked to from the examples - if you wanted to add some circles at specific lat and long on the graph, I thought you could add:

geo_y = [33.7]

geo_x = [84.3]

circle(geo_x, geo_y, radius=0.1, radius_units=“data”, color=“#3288BD”, tools=“”)

(again I was doing a usa plot)

into the plot - but what that did instead it created two plots - one of the usa, then a plot just below it on the produced webpage. - Is there a way to get them to layer on top of each other, so the geo plot is the base - then the circle is overlayed on it?

···

On Monday, January 13, 2014 2:40:42 PM UTC-5, Jared Thompson wrote:

Is there a way to not force it to the five colors, but to choose a gradient, then the scores can just go along the color line from 100 to 0 and choose the appropriate color?

Also, wouldn’t this then allow for a key to be place in the plot - is that possible?

3

Jared,

Add a call to hold() before the render commands, and they will all render to a single plot.

You are correct that there is no colorbar type legend yet. We will be releasing 0.4 in a few weeks and I expect more annotations/legends will show up in 0.5, together with an improved layout system.

You can definitely do a linear color mapping between fixed points, though there is not one built yet either on the python or the js side (pull requests welcome!) but that should be added fairly soon as well.

Bryan

···

On Jan 13, 2014, at 3:17 PM, Jared Thompson <[email protected]> wrote:

On Monday, January 13, 2014 2:40:42 PM UTC-5, Jared Thompson wrote:

Is there a way to not force it to the five colors, but to choose a gradient, then the scores can just go along the color line from 100 to 0 and choose the appropriate color?
Also, wouldn't this then allow for a key to be place in the plot - is that possible?

Looking through the source code I believe I am just seeing that keys are not yet implemented.
So I will ask another question. In the geo plot I linked to from the examples - if you wanted to add some circles at specific lat and long on the graph, I thought you could add:

geo_y = [33.7]
geo_x = [84.3]
circle(geo_x, geo_y, radius=0.1, radius_units="data", color="#3288BD", tools="")

(again I was doing a usa plot)

into the plot - but what that did instead it created two plots - one of the usa, then a plot just below it on the produced webpage. - Is there a way to get them to layer on top of each other, so the geo plot is the base - then the circle is overlayed on it?

--
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/671ee9e9-92ab-485a-95c4-c897756c7f03%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.

4

Add a call to hold() before the render commands, and they will all render to a single plot.
Should this do it?

hold()

show()

I will poke around to see if there is something additional I need to do as I still get two plots with that.

(pull requests welcome!) but that should be added fairly soon as well.

I would love to do that - though I think that is still in the future.

Thanks for your help.

···

On Monday, January 13, 2014 4:25:36 PM UTC-5, Bryan Van de ven wrote:

5

Jared,

Yes, typically you would do a hold() / show() pair.

If you are generating static HTML, with output_file(), then show() will also typically try to open a browser. If you just want to write the file you can use session().save() instead.

If you are in the ipython notebook, output_notebook(), then show() is probably always what you want.

Bryan

···

On Jan 13, 2014, at 3:39 PM, Jared Thompson <[email protected]> wrote:

On Monday, January 13, 2014 4:25:36 PM UTC-5, Bryan Van de ven wrote:
Add a call to hold() before the render commands, and they will all render to a single plot.
Should this do it?
hold()
show()

I will poke around to see if there is something additional I need to do as I still get two plots with that.

(pull requests welcome!) but that should be added fairly soon as well.

I would love to do that - though I think that is still in the future.
Thanks for your help.

--
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/0524c679-8463-43a1-bab7-2a0a50a97fa8%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.

6

Yeah, I just went into the examples directory and it worked right away on the provided examples, so I am probably just doing something funny in my customized one, so I will just have to play with it and move stuff around.

Thanks again for the additional insights.

···

On Monday, January 13, 2014 4:44:32 PM UTC-5, Bryan Van de ven wrote:

Jared,

Yes, typically you would do a hold() / show() pair.

7

Bryan,

I was wondering if you could help me understand how I should be thinking about this for bokeh, because what I am trying is not correct.

In the Texas example or my customized U.S.A. map example, it is working as long as I use:

colors = [“#F1EEF6”, “#D4B9DA”, “#C994C7”, “#DF65B0”, “#DD1C77”, “#980043”]

for the colors - but as stated the desire was to do a gradient, and I was brainstorming and testing that I should be able to do one color, but change the ‘fill_alpha’ state by state, that would give me the natural gradient across states for the data set.

I modified the fill color to just be black and set the fill_alpha to refer to a list of (for example): [0.15,0.23 … etc ]

county_colors =

patches(states_xs, states_ys, fill_color = “#000000”, fill_alpha=county_colors,

line_color=“white”, line_width=0.5, title=“tests”)

so the county_colors instead of being a list of colors, it is a list of fill_alpha settings.

But the issue is, each region’s fill alpha is solid black, instead of varying as it should with the fill_alpha=county_colors and the set of floats in there.

Should this approach work, or does bokeh not currently support the fill_alpha varying per region like that?

···

On Monday, January 13, 2014 4:25:36 PM UTC-5, Bryan Van de ven wrote:

You can definitely do a linear color mapping between fixed points

8

Jared,

Looks like you have found a bug, and that alphas are not treated correctly when there is a vector of them. I have added an issue you can follow:

  alpha not treated correctly as a data spec · Issue #271 · bokeh/bokeh · GitHub

This should be fixed int the upcoming 0.4 release.

Bryan

···

On Jan 14, 2014, at 9:06 AM, Jared Thompson <[email protected]> wrote:

On Monday, January 13, 2014 4:25:36 PM UTC-5, Bryan Van de ven wrote:
You can definitely do a linear color mapping between fixed points

Bryan,
I was wondering if you could help me understand how I should be thinking about this for bokeh, because what I am trying is not correct.
In the Texas example or my customized U.S.A. map example, it is working as long as I use:

colors = ["#F1EEF6", "#D4B9DA", "#C994C7", "#DF65B0", "#DD1C77", "#980043"]

for the colors - but as stated the desire was to do a gradient, and I was brainstorming and testing that I should be able to do one color, but change the 'fill_alpha' state by state, that would give me the natural gradient across states for the data set.

I modified the fill color to just be black and set the fill_alpha to refer to a list of (for example): [0.15,0.23 ... etc ]

county_colors =

patches(states_xs, states_ys, fill_color = "#000000", fill_alpha=county_colors,
        line_color="white", line_width=0.5, title="tests")

so the county_colors instead of being a list of colors, it is a list of fill_alpha settings.

But the issue is, each region's fill alpha is solid black, instead of varying as it should with the fill_alpha=county_colors and the set of floats in there.

Should this approach work, or does bokeh not currently support the fill_alpha varying per region like that?

--
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/10a057b7-5ace-44bc-8f23-1347f9074775%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.

9

Jared,

FYI this is fixed on GH master. If you want to try out GH let me know if you need any help installing, etc. Otherwise this will be in 0.4 released later this month. Thanks for the catch!

Bryan

···

On Jan 14, 2014, at 11:42 AM, Bryan Van de Ven <[email protected]> wrote:

Jared,

Looks like you have found a bug, and that alphas are not treated correctly when there is a vector of them. I have added an issue you can follow:

  alpha not treated correctly as a data spec · Issue #271 · bokeh/bokeh · GitHub

This should be fixed int the upcoming 0.4 release.

Bryan

On Jan 14, 2014, at 9:06 AM, Jared Thompson <[email protected]> wrote:

On Monday, January 13, 2014 4:25:36 PM UTC-5, Bryan Van de ven wrote:
You can definitely do a linear color mapping between fixed points

Bryan,
I was wondering if you could help me understand how I should be thinking about this for bokeh, because what I am trying is not correct.
In the Texas example or my customized U.S.A. map example, it is working as long as I use:

colors = ["#F1EEF6", "#D4B9DA", "#C994C7", "#DF65B0", "#DD1C77", "#980043"]

for the colors - but as stated the desire was to do a gradient, and I was brainstorming and testing that I should be able to do one color, but change the 'fill_alpha' state by state, that would give me the natural gradient across states for the data set.

I modified the fill color to just be black and set the fill_alpha to refer to a list of (for example): [0.15,0.23 ... etc ]

county_colors =

patches(states_xs, states_ys, fill_color = "#000000", fill_alpha=county_colors,
       line_color="white", line_width=0.5, title="tests")

so the county_colors instead of being a list of colors, it is a list of fill_alpha settings.

But the issue is, each region's fill alpha is solid black, instead of varying as it should with the fill_alpha=county_colors and the set of floats in there.

Should this approach work, or does bokeh not currently support the fill_alpha varying per region like that?

--
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/10a057b7-5ace-44bc-8f23-1347f9074775%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.

10

Bryan,
the part that was fixed - was that referring to the

hold() issue

or the

fill_alpha() not being able to take in a list to vary its values over the dataset?

Yes, I would love to test this - is there a guide somewhere on this? (running anaconda, so I am assuming I would create another environment)

···

On Tuesday, January 14, 2014 3:04:06 PM UTC-5, Bryan Van de ven wrote:

Jared,

FYI this is fixed on GH master. If you want to try out GH let me know if you need any help installing, etc. Otherwise this will be in 0.4 released later this month. Thanks for the catch!

Bryan

On Jan 14, 2014, at 11:42 AM, Bryan Van de Ven [email protected] wrote:

Jared,

Looks like you have found a bug, and that alphas are not treated correctly when there is a vector of them. I have added an issue you can follow:

    [https://github.com/ContinuumIO/bokeh/issues/271](https://github.com/ContinuumIO/bokeh/issues/271)

This should be fixed int the upcoming 0.4 release.

Bryan

On Jan 14, 2014, at 9:06 AM, Jared Thompson [email protected] wrote:

On Monday, January 13, 2014 4:25:36 PM UTC-5, Bryan Van de ven wrote:

You can definitely do a linear color mapping between fixed points

Bryan,

I was wondering if you could help me understand how I should be thinking about this for bokeh, because what I am trying is not correct.

In the Texas example or my customized U.S.A. map example, it is working as long as I use:

colors = [“#F1EEF6”, “#D4B9DA”, “#C994C7”, “#DF65B0”, “#DD1C77”, “#980043”]

for the colors - but as stated the desire was to do a gradient, and I was brainstorming and testing that I should be able to do one color, but change the ‘fill_alpha’ state by state, that would give me the natural gradient across states for the data set.

I modified the fill color to just be black and set the fill_alpha to refer to a list of (for example): [0.15,0.23 … etc ]

county_colors =

patches(states_xs, states_ys, fill_color = “#000000”, fill_alpha=county_colors,

   line_color="white", line_width=0.5, title="tests")

so the county_colors instead of being a list of colors, it is a list of fill_alpha settings.

But the issue is, each region’s fill alpha is solid black, instead of varying as it should with the fill_alpha=county_colors and the set of floats in there.

Should this approach work, or does bokeh not currently support the fill_alpha varying per region like that?

–
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/10a057b7-5ace-44bc-8f23-1347f9074775%40continuum.io.

For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out.

11

Jared,

Sorry, that was referring to eh fill_alpha not getting vectorized. I am looking at the other issue now.

Bryan

···

On Jan 14, 2014, at 3:00 PM, Jared Thompson <[email protected]> wrote:

Bryan,
the part that was fixed - was that referring to the
hold() issue
or the
fill_alpha() not being able to take in a list to vary its values over the dataset?

Yes, I would love to test this - is there a guide somewhere on this? (running anaconda, so I am assuming I would create another environment)

On Tuesday, January 14, 2014 3:04:06 PM UTC-5, Bryan Van de ven wrote:
Jared,

FYI this is fixed on GH master. If you want to try out GH let me know if you need any help installing, etc. Otherwise this will be in 0.4 released later this month. Thanks for the catch!

Bryan

On Jan 14, 2014, at 11:42 AM, Bryan Van de Ven <[email protected]> wrote:

> Jared,
>
> Looks like you have found a bug, and that alphas are not treated correctly when there is a vector of them. I have added an issue you can follow:
>
> alpha not treated correctly as a data spec · Issue #271 · bokeh/bokeh · GitHub
>
> This should be fixed int the upcoming 0.4 release.
>
> Bryan
>
> On Jan 14, 2014, at 9:06 AM, Jared Thompson <[email protected]> wrote:
>
>>
>>
>> On Monday, January 13, 2014 4:25:36 PM UTC-5, Bryan Van de ven wrote:
>> You can definitely do a linear color mapping between fixed points
>>
>> Bryan,
>> I was wondering if you could help me understand how I should be thinking about this for bokeh, because what I am trying is not correct.
>> In the Texas example or my customized U.S.A. map example, it is working as long as I use:
>>
>> colors = ["#F1EEF6", "#D4B9DA", "#C994C7", "#DF65B0", "#DD1C77", "#980043"]
>>
>> for the colors - but as stated the desire was to do a gradient, and I was brainstorming and testing that I should be able to do one color, but change the 'fill_alpha' state by state, that would give me the natural gradient across states for the data set.
>>
>>
>> I modified the fill color to just be black and set the fill_alpha to refer to a list of (for example): [0.15,0.23 ... etc ]
>>
>>
>> county_colors =
>>
>> patches(states_xs, states_ys, fill_color = "#000000", fill_alpha=county_colors,
>> line_color="white", line_width=0.5, title="tests")
>>
>> so the county_colors instead of being a list of colors, it is a list of fill_alpha settings.
>>
>> But the issue is, each region's fill alpha is solid black, instead of varying as it should with the fill_alpha=county_colors and the set of floats in there.
>>
>> Should this approach work, or does bokeh not currently support the fill_alpha varying per region like that?
>>
>> --
>> 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/10a057b7-5ace-44bc-8f23-1347f9074775%40continuum.io\.
>> For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.
>

--
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/a977cc0e-8914-4d5a-8391-3fa4e2f2e272%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.

12

Jared,

the other issue is fixed now in master as well.

Bryan

···

On Jan 14, 2014, at 3:00 PM, Jared Thompson <[email protected]> wrote:

Bryan,
the part that was fixed - was that referring to the
hold() issue
or the
fill_alpha() not being able to take in a list to vary its values over the dataset?

Yes, I would love to test this - is there a guide somewhere on this? (running anaconda, so I am assuming I would create another environment)

On Tuesday, January 14, 2014 3:04:06 PM UTC-5, Bryan Van de ven wrote:
Jared,

FYI this is fixed on GH master. If you want to try out GH let me know if you need any help installing, etc. Otherwise this will be in 0.4 released later this month. Thanks for the catch!

Bryan

On Jan 14, 2014, at 11:42 AM, Bryan Van de Ven <[email protected]> wrote:

> Jared,
>
> Looks like you have found a bug, and that alphas are not treated correctly when there is a vector of them. I have added an issue you can follow:
>
> alpha not treated correctly as a data spec · Issue #271 · bokeh/bokeh · GitHub
>
> This should be fixed int the upcoming 0.4 release.
>
> Bryan
>
> On Jan 14, 2014, at 9:06 AM, Jared Thompson <[email protected]> wrote:
>
>>
>>
>> On Monday, January 13, 2014 4:25:36 PM UTC-5, Bryan Van de ven wrote:
>> You can definitely do a linear color mapping between fixed points
>>
>> Bryan,
>> I was wondering if you could help me understand how I should be thinking about this for bokeh, because what I am trying is not correct.
>> In the Texas example or my customized U.S.A. map example, it is working as long as I use:
>>
>> colors = ["#F1EEF6", "#D4B9DA", "#C994C7", "#DF65B0", "#DD1C77", "#980043"]
>>
>> for the colors - but as stated the desire was to do a gradient, and I was brainstorming and testing that I should be able to do one color, but change the 'fill_alpha' state by state, that would give me the natural gradient across states for the data set.
>>
>>
>> I modified the fill color to just be black and set the fill_alpha to refer to a list of (for example): [0.15,0.23 ... etc ]
>>
>>
>> county_colors =
>>
>> patches(states_xs, states_ys, fill_color = "#000000", fill_alpha=county_colors,
>> line_color="white", line_width=0.5, title="tests")
>>
>> so the county_colors instead of being a list of colors, it is a list of fill_alpha settings.
>>
>> But the issue is, each region's fill alpha is solid black, instead of varying as it should with the fill_alpha=county_colors and the set of floats in there.
>>
>> Should this approach work, or does bokeh not currently support the fill_alpha varying per region like that?
>>
>> --
>> 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/10a057b7-5ace-44bc-8f23-1347f9074775%40continuum.io\.
>> For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.
>

--
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/a977cc0e-8914-4d5a-8391-3fa4e2f2e272%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.

13

Bryan,

I went to test this and got an error. In just doing the texas ipython notebook (no modifications just the one in the example notebook)

I get this error after executing the “show()” cell in the notebook:

Plots

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

See your browser Javascript console for more details.

Perhaps I have done something incorrect, but to do this I did:

cloned the repo:

git clone GitHub - bokeh/bokeh: Interactive Data Visualization in the browser, from Python

conda create -n bokeh anaconda
conda remove bokeh

then I went into the git cloned repo folder of bokeh

source activate bokeh
python setup.py install

then ran the texas example and received:

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

···

On Tuesday, January 14, 2014 4:45:52 PM UTC-5, Bryan Van de ven wrote:

Jared,

the other issue is fixed now in master as well.

Bryan

14

Just correcting the steps I took:

Perhaps I have done something incorrect, but to do this I did:

cloned the repo:

git clone https://github.com/ContinuumIO/bokeh.git

conda create -n bokeh anaconda

then I went into the git cloned repo folder of bokeh

source activate bokeh

conda remove bokeh

python setup.py install

then ran the texas example and received:

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

···

On Wednesday, January 15, 2014 7:54:09 AM UTC-5, Jared Thompson wrote:

On Tuesday, January 14, 2014 4:45:52 PM UTC-5, Bryan Van de ven wrote:

Jared,

the other issue is fixed now in master as well.

Bryan

Bryan,

I went to test this and got an error. In just doing the texas ipython notebook (no modifications just the one in the example notebook)

I get this error after executing the “show()” cell in the notebook:

Plots

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

See your browser Javascript console for more details.

Perhaps I have done something incorrect, but to do this I did:

cloned the repo:

git clone https://github.com/ContinuumIO/bokeh.git

conda create -n bokeh anaconda
conda remove bokeh

then I went into the git cloned repo folder of bokeh

source activate bokeh
python setup.py install

then ran the texas example and received:

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

15

Jared,

Ah no but you need to also rebuild bokehjs. By default the last bokehjs release gets installed but some things have changed since then and so bokehjs will also need to get built. I am on my phone atm I will send better instructions for rebuilding bokehjs when I get to a computer. Or you can check out the dev guide on bokeh.pydata.org in the mean time.

···

On Jan 15, 2014, at 6:54, Jared Thompson [email protected] wrote:

On Tuesday, January 14, 2014 4:45:52 PM UTC-5, Bryan Van de ven wrote:

Jared,

the other issue is fixed now in master as well.

Bryan

Bryan,

I went to test this and got an error. In just doing the texas ipython notebook (no modifications just the one in the example notebook)

I get this error after executing the “show()” cell in the notebook:

Plots

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

See your browser Javascript console for more details.

Perhaps I have done something incorrect, but to do this I did:

cloned the repo:

git clone https://github.com/ContinuumIO/bokeh.git

conda create -n bokeh anaconda
conda remove bokeh

then I went into the git cloned repo folder of bokeh

source activate bokeh
python setup.py install

then ran the texas example and received:

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

–

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/72b5f0d9-2535-415e-b177-b1073df75859%40continuum.io.

For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out.

16

Thanks for the link,
I will follow the instructions on:

http://bokeh.pydata.org/dev_guide.html#installation-for-developers

···

On Wednesday, January 15, 2014 8:11:16 AM UTC-5, Bryan Van de ven wrote:

Jared,

Ah no but you need to also rebuild bokehjs. By default the last bokehjs release gets installed but some things have changed since then and so bokehjs will also need to get built. I am on my phone atm I will send better instructions for rebuilding bokehjs when I get to a computer. Or you can check out the dev guide on bokeh.pydata.org in the mean time.

On Jan 15, 2014, at 6:54, Jared Thompson [email protected] wrote:

On Tuesday, January 14, 2014 4:45:52 PM UTC-5, Bryan Van de ven wrote:

Jared,

the other issue is fixed now in master as well.

Bryan

Bryan,

I went to test this and got an error. In just doing the texas ipython notebook (no modifications just the one in the example notebook)

I get this error after executing the “show()” cell in the notebook:

Plots

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

See your browser Javascript console for more details.

Perhaps I have done something incorrect, but to do this I did:

cloned the repo:

git clone https://github.com/ContinuumIO/bokeh.git

conda create -n bokeh anaconda
conda remove bokeh

then I went into the git cloned repo folder of bokeh

source activate bokeh
python setup.py install

then ran the texas example and received:

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

–

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/72b5f0d9-2535-415e-b177-b1073df75859%40continuum.io.

For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out.

17

Bryan,
thanks for all the help - it works now and I can see the fixes.

I had never used npm, so it just took a bit to figure out you have to run the commands from http://bokeh.pydata.org/dev_guide.html#installation-for-developers

within the bokehjs folder.

Now everything is running and I can fully test things now.

···

On Wednesday, January 15, 2014 8:49:44 AM UTC-5, Jared Thompson wrote:

Thanks for the link,
I will follow the instructions on:

http://bokeh.pydata.org/dev_guide.html#installation-for-developers

On Wednesday, January 15, 2014 8:11:16 AM UTC-5, Bryan Van de ven wrote:

Jared,

Ah no but you need to also rebuild bokehjs. By default the last bokehjs release gets installed but some things have changed since then and so bokehjs will also need to get built. I am on my phone atm I will send better instructions for rebuilding bokehjs when I get to a computer. Or you can check out the dev guide on bokeh.pydata.org in the mean time.

On Jan 15, 2014, at 6:54, Jared Thompson [email protected] wrote:

On Tuesday, January 14, 2014 4:45:52 PM UTC-5, Bryan Van de ven wrote:

Jared,

the other issue is fixed now in master as well.

Bryan

Bryan,

I went to test this and got an error. In just doing the texas ipython notebook (no modifications just the one in the example notebook)

I get this error after executing the “show()” cell in the notebook:

Plots

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

See your browser Javascript console for more details.

Perhaps I have done something incorrect, but to do this I did:

cloned the repo:

git clone https://github.com/ContinuumIO/bokeh.git

conda create -n bokeh anaconda
conda remove bokeh

then I went into the git cloned repo folder of bokeh

source activate bokeh
python setup.py install

then ran the texas example and received:

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

–

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/72b5f0d9-2535-415e-b177-b1073df75859%40continuum.io.

For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out.

18

Jared,

great, thanks for the feedback I have made a ticket for clarifying the dev docs:

  improve dev docs about how to build bokehjs · Issue #283 · bokeh/bokeh · GitHub

Bryan

···

On Jan 15, 2014, at 4:37 PM, Jared Thompson <[email protected]> wrote:

Bryan,
thanks for all the help - it works now and I can see the fixes.

I had never used npm, so it just took a bit to figure out you have to run the commands from http://bokeh.pydata.org/dev_guide.html#installation-for-developers
within the bokehjs folder.

Now everything is running and I can fully test things now.

On Wednesday, January 15, 2014 8:49:44 AM UTC-5, Jared Thompson wrote:
Thanks for the link,
I will follow the instructions on:
http://bokeh.pydata.org/dev_guide.html#installation-for-developers

On Wednesday, January 15, 2014 8:11:16 AM UTC-5, Bryan Van de ven wrote:
Jared,

Ah no but you need to also rebuild bokehjs. By default the last bokehjs release gets installed but some things have changed since then and so bokehjs will also need to get built. I am on my phone atm I will send better instructions for rebuilding bokehjs when I get to a computer. Or you can check out the dev guide on bokeh.pydata.org in the mean time.

On Jan 15, 2014, at 6:54, Jared Thompson <[email protected]> wrote:

On Tuesday, January 14, 2014 4:45:52 PM UTC-5, Bryan Van de ven wrote:
Jared,

the other issue is fixed now in master as well.

Bryan

Bryan,
I went to test this and got an error. In just doing the texas ipython notebook (no modifications just the one in the example notebook)

I get this error after executing the "show()" cell in the notebook:

Plots
Javascript error adding output!
./base: Unknown Collection WheelZoomTool
See your browser Javascript console for more details.

Perhaps I have done something incorrect, but to do this I did:

cloned the repo:
git clone GitHub - bokeh/bokeh: Interactive Data Visualization in the browser, from Python

conda create -n bokeh anaconda
conda remove bokeh

then I went into the git cloned repo folder of bokeh
source activate bokeh
python setup.py install

then ran the texas example and received:

Javascript error adding output!
./base: Unknown Collection WheelZoomTool

--
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/72b5f0d9-2535-415e-b177-b1073df75859%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.

--
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/e527d7a6-ab58-4fec-b0ef-6ec41e00f7c5%40continuum.io\.
For more options, visit https://groups.google.com/a/continuum.io/groups/opt_out\.