Not showing all rows in response

[Hrishikesh007788]

I used a csv file having 9000+ rows. When I ask for a response, it doesn't show all rows and uses dots. What should I do if I want to see all the rows?

[gventuri]

Hi @Hrishikesh007788 thanks for asking. Could you copy/paste the code you are using? Maybe it's filtering out some rows as part of the query. Let me know!

[Hrishikesh007788]

Here is the log of what's happening the the background. I am trying to see the list of customers but its giving me the answer in this form. The output has 521 rows but its not showing everything. Is there a way to display every row? Since I am displaying the "Answer:" as it is in the frontend, this type of output is not much useful.

2023-08-08 17:55:15 [INFO] Question: Customers generating revenue more than 10000 in FY20 Q3
2023-08-08 17:55:15 [INFO] Running PandasAI with openai LLM...
2023-08-08 17:55:15 [INFO] Prompt ID: 799acb40-b36a-4e02-b2b3-c6d0c185164b
2023-08-08 17:55:15 [INFO] Using cached response
2023-08-08 17:55:15 [INFO]
Code running:

df[df['FY20_Q3_REVENUE'] > 10000]['Customer']

2023-08-08 17:55:15 [INFO] Answer: 49 Customer 50
57 Customer 57
84 Customer 82
85 Customer 83
98 Customer 93
...
8990 Customer 8467
8998 Customer 8475
9010 Customer 8487
9021 Customer 8496
9050 Customer 8525
Name: Customer, Length: 521, dtype: object

[mspronesti]

To solve this one simply needs to add on top pd.set_option('display.max_rows', None). @gventuri Do you think it is worth it to make this a standard behavior ?

[gventuri]

@mspronesti yep, 100% agree. Just wondering at which point we'd better change the behavior. As we initialize PandasAI?

[mspronesti]

We might even just put the following two lines on top of the generated code, after polishing it

pd.set_option('display.max_columns', None)
pd.set_option('dipslay.max_rows', None)

Not sure this is a desired behavior though. Is it helpful to visualize a big dataframe in the console?

[Hrishikesh007788]

I am displaying it in the frontend. Is there a way to make it more presentable? Dataframe looks too raw in the frontend