Colombia municipal data count mismatch

[jsvine]

Data parsed from Colombia's latest municipality-level PDF provides these totals:

• Confirmados por laboratorio: 1,050
• Confirmados por clínica: 17,115
• Sospechosos: 2,132

But the sum of these columns from the data I've parsed provides these totals:

• Confirmados por laboratorio: 1,049 (-1)
• Confirmados por clínica: 17,112 (-3)
• Sospechosos: 2,117 (-15)

Am I missing one/several municipalities? Are the official totals wrong? Do they include unlisted municipalities? I've scoured the two files, but can't find the source of the discrepancy. Any ideas?

[chendaniely]

@bryanleroylewis you mentioned [1] that we (@ndssl/zika) were tapped for Colombia data. Any thoughts about this data quality issue?

The frequency counts seem small enough (relatively) that we can start cleaning it up for the CDC now, but we should still probably address this issue eventually

[1] https://github.com/ndssl/zika_data_to_cdc/pull/1