[BUG] Unmatched values in comparing q78 if problematic_val is 0.0

[yinqingh]

Describe the bug
In comparing q78 problematic values with SF30K datasets, I observed unmatched values error in the results but two rows are identical

[2025-04-01T10:01:51.743Z] Collected 100 rows in 2.24267315864563 seconds
[2025-04-01T10:01:51.743Z] Row 96: 
[2025-04-01T10:01:51.743Z] [7900, 0.0, 44, Decimal('99.00'), Decimal('15.92'), 19788, Decimal('16291.44'), Decimal('19716.60')]
[2025-04-01T10:01:51.743Z] [7900, 0.0, 44, Decimal('99.00'), Decimal('15.92'), 19788, Decimal('16291.44'), Decimal('19716.60')]
[2025-04-01T10:01:51.743Z] 
[2025-04-01T10:01:51.743Z] Processed 100 rows
[2025-04-01T10:01:51.743Z] There were 1 errors
[2025-04-01T10:01:51.743Z] === Unmatch Queries: ['query78'] ===

Based on the q78 compare function, the problematic_val_eq will be false if problematic_val_row1 and problematic_val_row2 is 0.0, which is unexpected. This should be the root cause of this issue.

# this value could be none in some rows
if all([problematic_val_row1, problematic_val_row2]):
    # this value is rounded to its pencentile: round(ss_qty/(coalesce(ws_qty,0)+coalesce(cs_qty,0)),2)
    # so we allow the diff <= 0.01 + default epsilon 0.00001
    problematic_val_eq = abs(problematic_val_row1 - problematic_val_row2) <= 0.01001
elif problematic_val_row1 == None and problematic_val_row2 == None:
    problematic_val_eq = True
else:
    problematic_val_eq = False
return problematic_val_eq and all([compare(lhs, rhs, epsilon) for lhs, rhs in zip(row1, row2)])

Debug log:

[2025-04-02T02:35:20.693Z] problematic_val_row1: 0.0
[2025-04-02T02:35:20.693Z] problematic_val_row2: 0.0
[2025-04-02T02:35:20.693Z] problematic_val_eq: False
[2025-04-02T02:35:20.693Z] 7900 7900 True
[2025-04-02T02:35:20.693Z] 44 44 True
[2025-04-02T02:35:20.693Z] 99.00 99.00 True
[2025-04-02T02:35:20.693Z] 15.92 15.92 True
[2025-04-02T02:35:20.693Z] 19788 19788 True
[2025-04-02T02:35:20.693Z] 16291.44 16291.44 True
[2025-04-02T02:35:20.693Z] 19716.60 19716.60 True
[2025-04-02T02:35:20.693Z] True
[2025-04-02T02:35:20.693Z] epsilon: 1e-05
[2025-04-02T02:35:20.693Z] is_q78: True
[2025-04-02T02:35:20.693Z] q78_problematic_col: 2
[2025-04-02T02:35:20.693Z] Row 96: 
[2025-04-02T02:35:20.693Z] [7900, 0.0, 44, Decimal('99.00'), Decimal('15.92'), 19788, Decimal('16291.44'), Decimal('19716.60')]
[2025-04-02T02:35:20.693Z] [7900, 0.0, 44, Decimal('99.00'), Decimal('15.92'), 19788, Decimal('16291.44'), Decimal('19716.60')]

Steps/Code to reproduce bug
Please provide a list of steps or a code sample to reproduce the issue.
Avoid posting private or sensitive data.

Expected behavior
A clear and concise description of what you expected to happen.

Environment details (please complete the following information)

• Environment
• Configuration settings related to the issue

Additional context
Add any other context about the problem here.