A missing value marks an observation that is unavailable, inapplicable, or not reported. Raw data may encode it as the string "Missing"; replacing that marker with None, an empty string, or pandas NaN makes its status explicit. section 21.4.1 section 24.3.1
Handle missing values deliberately when cleaning data:
- Convert "Missing" while reading the file, or guard every calculation so it skips that string. section 21.4.1
- Treat missing heat-index fields as distinct from percentage fields: both may be parsed as strings, but percentages require numeric translation before averaging. section 21.4.1 section 21.6
- Standardize varied markers such as NA, N/A, ?, and - by supplying them together as pandas read_csv() na_values. section 24.3.1
Converting a raw marker
Use `temp = pd.read_csv("temp_data_01.csv", na_values=['Missing'])`. A field containing `Missing` then appears as `NaN`, while ordinary temperature and count values remain numeric. The resulting missing value can be excluded or handled explicitly during later calculations. section 24.3.1
[None, 'Illinois', 17, 'Jan 01, 1979', '1979/01/01', 17.48, 994, 6, 30.5, 2.89, 994, -13.6, 15.8, 'Missing', 0, 'Missing', 'Missing', '0.00%']
Inspect a representative weather row to see missing heat-index values beside numeric temperatures, counts, and a percentage marker. section 21.4.1
