You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
multiperiod balance report should be more like bs in showing empty columns ? #1396
changed the title [-]multiperiod balance report should be more respectful of a specified report period[/-][+]multiperiod balance report should be more like bs in showing empty columns ?[/+]on Nov 17, 2020
A question regarding fixing this issue:
Making this change obviously adds a bunch of extra column an the start and end in some cases: for example, test 2 in intervals.test.
hledger -f balance-multicol.journal balance -p 'monthly in 2013' --no-total
>>>
Balance changes in 2013Q1:
|| Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
=================++============================================================
assets || 0 1 0 0 0 0 0 0 0 0 0 0
assets:cash || 0 1 0 0 0 0 0 0 0 0 0 0
assets:checking || 0 0 1 0 0 0 0 0 0 0 0 0
This is even if the most recent posting is in March 2013, as in this case. Should we automatically intersect the report span with the journal span? This would leave us with no way of getting reports beyond the span of the journal without adding dummy transactions, but might be a reasonable compromise on usability.
A noteable thing about making this change is that the report title will change, which may actually be a good thing. In the above example it would change to:
hledger -f balance-multicol.journal balance -p 'monthly in 2013' --no-total
>>>
Balance changes in 2013Q1:
|| Jan Feb Mar
=================++===============
assets || 0 1 0
assets:cash || 0 1 0
assets:checking || 0 0 1
A side note: The report title is probably better off mentioning the dates, rather than "Balance changes in 2020Q1", since the meaning of "Q1", "Q2", etc., aren't fixed. For instance, in Australia, Q1 refers to the period July 1–September 30, whereas in India, it is usually the period April 1–June 30.
Activity
simonmichael commentedon Nov 17, 2020
Ah it's because
balancealso hides leading/trailing all-zero columns by default, so you must add -E/--empty:Should they be more consistent ? Which behaviour is more robust and useful with typical real-world data ?
[-]multiperiod balance report should be more respectful of a specified report period[/-][+]multiperiod balance report should be more like bs in showing empty columns ?[/+]Xitian9 commentedon Dec 4, 2020
I definitely would agree with this change; it has annoyed me mildly in the past. It's a simple change; I'll look into it when I get some time.
Xitian9 commentedon Dec 4, 2020
Just linking to a previous discussion on the topic of empty columns in #1269
Xitian9 commentedon Dec 14, 2020
A question regarding fixing this issue:
Making this change obviously adds a bunch of extra column an the start and end in some cases: for example, test 2 in intervals.test.
This is even if the most recent posting is in March 2013, as in this case. Should we automatically intersect the report span with the journal span? This would leave us with no way of getting reports beyond the span of the journal without adding dummy transactions, but might be a reasonable compromise on usability.
Xitian9 commentedon Dec 14, 2020
A noteable thing about making this change is that the report title will change, which may actually be a good thing. In the above example it would change to:
the-solipsist commentedon Jan 6, 2021
A side note: The report title is probably better off mentioning the dates, rather than "Balance changes in 2020Q1", since the meaning of "Q1", "Q2", etc., aren't fixed. For instance, in Australia, Q1 refers to the period July 1–September 30, whereas in India, it is usually the period April 1–June 30.
simonmichael commentedon Jan 6, 2021
That's interesting, thanks for mentioning.
Xitian9 commentedon Jan 7, 2021
Future considerations for localisation aside, I think this issue might be resolved now.