[R-SIG-Finance] as.journal returns subscript out of bounds

[Enrico Schumann]

On Fri, 11 Apr 2025, Arnaud Gaboury writes:

> Even in a new empty environment, anf after a rebuild of ally my data
> frames which bring me to the final ones, I still get this error:
> gabx using magnolia [R] btc_journal <- as.journal(btc_op)
> gabx using magnolia [R] View(btc_journal)
> Error in names[[i]] : subscript out of bounds
> gabx using magnolia [R] str(btc_journal)
> ‘journal’:	 78 transactions
>  $ instrument: chr [1:78] "BTCUSDC" "BTCUSDC" "BTCUSDC" "BTCUSDC" ...
>  $ timestamp : POSIXct[1:78], format: "2024-12-16 18:10:00" "2024-12-17
> 09:01:30" "2024-12-18 20:32:53" "2025-01-02 09:33:18" ...
>  $ amount    : num [1:78] 9.3372 0.0481 0.1232 -0.065 0.9656 ...
>  $ price     : num [1:78] 107012 107359 101423 96408 96606 ... 
>
>
>
> I am afraid I must learn to use traceback() as I can't see any obvious
> reasons for the error.

I cannot reproduce with recent Rs on GNU/Linux.

But I _can_ reproduce it on Windows with RStudio. And
the error is not thrown by `as.journal`, but when I try
to call `View`.  Which is RStudio's own variant of the
function.  What happens when you say

    utils::View(btc_journal)

?


    Enrico

> On Fri, 2025-04-11 at 07:40 -0500, Dirk Eddelbuettel wrote:
>> 
>> On 11 April 2025 at 14:16, Arnaud Gaboury wrote:
>> > I can't see any reason why the command returns an error.
>> 
>> It does not error for me:
>> 
>>   > class(btc_op)
>>   [1] "tbl_df"     "tbl"        "data.frame"
>>   > class(as.journal(btc_op))
>>   [1] "journal"
>>   > head(as.journal(btc_op))
>>      instrument            timestamp    amount     price
>>   1     BTCUSDC  2024-12-16 12:10:00   9.33721  107012.0
>>   2     BTCUSDC  2024-12-17 03:01:30   0.04811  107358.8
>>   3     BTCUSDC  2024-12-18 14:32:53   0.12324  101422.9
>>   4     BTCUSDC  2025-01-02 03:33:18  -0.06504   96408.2
>>   5     BTCUSDC  2025-01-02 04:15:12   0.96557   96606.1
>>   6     BTCUSDC  2025-01-14 05:34:28   2.42992   96526.3
>>   
>>   6 transactions  
>>   > 
>> 
>> So maybe your local setup is borked. I am 'CRAN-current' here and
>> used the
>> `structure()` command you posted.
>> 
>> You may want to learn about traceback(), and things like browser(),
>> to debug
>> functions.  It's all open source so you can play with it.
>> 
>> Dirk

-- 
Enrico Schumann
Lucerne, Switzerland
https://enricoschumann.net