Plotly Not Working in CODEBK?

Aleniles

Hello, trying to use plotly to plot this simple dataframe (dfplot):

     index    Strat1
237 2022-11-29 -0.159187
238 2022-11-30 3.094787
239 2022-12-01 -0.086763
240 2022-12-02 -0.000000
241 2022-12-05 -0.000000

Using this code:

import plotly.express as px
fig = px.line(dfplot, x="Date", y="Daily Return", title='Daily Return Historical Chart')
fig.show()

But I get the error:

AttributeError: module 'dask.array' has no attribute 'lib'

AttributeError                            Traceback (most recent call last)
/tmp/ipykernel_200/32131141.py in <module>
----> 1 import plotly.express as px

Is there any dependencies not working properly ?

wasin.w

Hello @Aleniles

Thank you for the information. I checked my Eikon (version 4.0.60018, not the latest). The CodeBook App on my Eikon is version 3.0 (March 2022) with the following dependencies versions:

• JupyterLab 3.2.4
• Python 3.8.13
• Plotly 5.4.0
• Dask 2021.11.2

You can check the version from the CodeBook "_Changelog.md" file in the main folder as follows:

I highly recommend you contact the Eikon support team directly to help you with this CodeBook app and libraries issue.

In the meantime, can you try the CodeBook app on the Eikon Web version? Please find more detail in this Use CodeBook on Web Browser article.

wasin.w

Hello @Aleniles

Are you using the latest version of the Workspace desktop application?

I am testing the following code with CodeBook app on Workspace version 1.19.448 and it works fine.

Firstly, I create the dataframe object and set the "Date" column to be a normal column instead.

Code:

# Creates pandas DataFrame.
data = {'Strat1': [-0.159187,3.094787,-0.086763, -0.000000, -0.000000]}
dfplot= pd.DataFrame(data, index=['2022-11-29','2022-11-30','2022-12-02','2022-12-05','2022-12-05'])
# Reset index to include Date as a data column
dfplot.reset_index(level=0, inplace=True)
dfplot.rename({'index':'Date'}, axis= 1, inplace= True)

Result:

Then I plot the graph with plotly:

fig = px.line(dfplot, x="Date", y="Strat1", title='Daily Return Historical Chart')
fig.show()

Result

wasin.w

Hello @Aleniles

You can check the version of Workspace from the Workspace desktop app --> Help (?) menu --> "About Refinitiv Workspace" button.

Aleniles

No, it still doesn't work because the error is directly on the "import plotly.express as px" instruction. That's why I am thinking about "dependencies" problems.

Btw I do not have the same view as you in the workspace. I am running CODEBOOK with Datastream data inside Refinitiv Desktop Eikon 4.0.54 with JupyterLab 1.2.7, plotly 4.8.1

. The python version is the one below.

Unfortunately I am not able to upload a picture due to IT restrictions but believe me: running a simple cell with the "import plotly.express as px" instruction gives the error I mentioned.

Python 3.8.13 | packaged by conda-forge | (default, Mar 25 2022, 06:04:10) 
Type 'copyright', 'credits' or 'license' for more information
IPython 7.31.0 -- An enhanced Interactive Python. Type '?' for help.

Aleniles

See this https://stackoverflow.com/questions/73099932/attributeerror-module-dask-array-has-no-attribute-lib

Aleniles

Many thanks.

So in the standalone version I do not have the changelog file either ... so I cannot check.

I tried to connect to the web version and here, indeed, everything works fine. So I suppose I have an outdated version of Codebook installed on my machine, I'll reach out the Refinitiv helpdesk.

One last question:

Do you know how or if it is possible to connect to external APIs using Codebook ?

I have tried this:

import yfinance as yf
Amazon = yf.Ticker("AMZN")
print(Amazon.info) 

But I am getting this:

HTTPSConnectionPool(host='finance.yahoo.com', port=443): Max retries exceeded with url: /quote/AMZN (Caused by ProxyError('Cannot connect to proxy.', OSError('Tunnel connection failed: 403 Forbidden')))

I'm wondering if there is a way to set up something (whitelisting servers, setting the SSL or Proxy properly etc) so that it could work.

wasin.w

Hello @Aleniles

Unfortunately, the CodeBook app allows connection within Refinitiv platforms (Eikon, Refinitiv Data Platform, DataStream, etc) only.
If your notebook app needs to connect to external services/APIs, you need to set up your local Jupyter Lab environment and run a notebook app in your local Jupyter instead. You can find more detail about how to set up Jupyter Lab from the following resources:

• Installing Jupyter
• Setup a Data Science Environment on your Computer
• How to Setup Your JupyterLab Project Environment

Caution!

Please contact your Refinitiv representative to confirm if you need a separate Datastream credential to run Datastream in your local environment.