Showing plots in Jupyter notebooks

[Paul Smart]

The plt.show() in this simple example works fine:

plt.plot(range(20), range(20))

plt.show()

print "hello world"

However, this example returns the error: AttributeError: 'AxesSubplot' object has no attribute 'show'

    #Plot results

    data=pd.DataFrame.from_records(analysis['analysis']['results'], index='key')

    plt = data.plot(kind = 'line',title=id['label']+" : "+name, xticks = xkeys)

    plt.set_xticklabels(xlabels, rotation='vertical');

    plt.show()

Can anyone help me to ensure the plot is displayed at the correct time in this second example please?

(otherwise it waits until the end of the cell output)

Thanks,

 Paul

[William Stein]

Could you provide a working example? Also, this seems like a pure
Jupyter notebook question, so you might get better help on their
mailing list.

William

>
> Thanks,
> Paul
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-cloud" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-cloud+...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-cloud/76cfa9da-1aaa-49ba-91c8-5cba1aaf85b9%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.



--

Best Regards,
William

Found/CEO of SageMath, Inc.

Professor of Mathematics

[Harald Schilly]

Hi, this is actually only a matplotlib question.

On Wed, Nov 25, 2015 at 5:59 PM, Paul Smart <paul....@datasift.com> wrote:
> #Plot results
>     data=pd.DataFrame.from_records(analysis['analysis']['results'],
> index='key')
>     plt = data.plot(kind = 'line',title=id['label']+" : "+name, xticks =
> xkeys)
>     plt.set_xticklabels(xlabels, rotation='vertical');
>     plt.show()

You are confused, because you are assigning the axes object to

​ the variable​

plt in line

​2

.

Here is a link to the api of matplotlib's axes:
http://matplotlib.org/1.4.3/api/axes_api.html

Here is a revised example that works out of the box, very similar to yours -- but I'm missing the data to reproduce it.

%matplotlib inline
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import string
data=pd.DataFrame([1,2,3,4,3,4,3,2,1])
ax = data.plot(kind = 'line',title="label")
ax.set_xticklabels(string.uppercase[:len(data)], rotation='vertical')
plt.show()

-- harald

[Paul Smart]

Thanks a million Harald - that works great.

This is what happen when non-developers like me copy and paste working code rather than actually understanding it!

[Harald Schilly]

On Thu, Nov 26, 2015 at 11:28 AM, Paul Smart <paul....@datasift.com> wrote:

This is what happen when non-developers like me copy and paste working code rather than actually understanding it!

Yes, the key idea is,that the variable name in itself doesn't convey​ any information. So, you have to do things like "type(plt)" to learn which type the object is, where the variable "plt" is pointing to. 

Then, you would have noticed that the return value of the pandas plot is an Axis object.

PS: In case you are lost and need the current figure or axis of matplotlib: it's plt.gcf() and plt.gca() if plt is this pyplot api you are canonically importing in the beginning.

-- harald