HEATMAPSTER — VIDEO EXERCISES

All data is fictional. HeatMapster_examples.xlsx contains the same original sample data as the supplied tutorial workbook, with an English file name.

H01: Ready_matrix, A1:E5. Two blank cells remain missing, not zero.
H02: Sales_records. North / Desk sums 1200 + 350 = 1550; Average = 775; Count = 2.
H03: Actual_budget. Sales / Q1: Current 760, Baseline 700, Difference 60; Percentage Change about 8.57%.
H04: Activity_4weeks.csv. Import as Raw Table, validate the complete file, and review the warning about distinct timestamps. Time & Calendar groups timestamps into time buckets; the warning is expected for this example. Day of Week × Hour, Monday, one hour, all years combined, weekends included: 1132 records, Monday at 09:00 = 48. An empty combination and a numeric zero are different display choices.
H05: Use the H02 heatmap. Export an exact-size 1920 × 1080 PNG with Advanced → Label font (DIP) = 24 for this small example, an A4 landscape PDF fitted to one page, and a widescreen PowerPoint with an editable heatmap. Optional title/summary slides are disabled in the video.
H06: Make a working copy of the workbook named Sales_refresh.xlsx in your own folder. Import Sales_records A1:C9 as Raw Table, build Region / Product / Sum / Sales, and save a .heatmap project. Change only C2 from 1200 to 1500 in that same workbook, save, and close the workbook in Excel to release its file lock before refreshing. The map remains 1550 until you choose Refresh Source, review the prepared update, and Apply Refresh. The new total is 1850 and overall sales 5170. Save the project again. The two reference CSVs are value checks, not two alternative source paths. Do not use a built-in sample for this exercise: it has no external file source.
H07: Correlation. Select all five numeric variables, including Constant. Pearson, Pairwise, minimum n = 3 is a demonstration setting, not a reliability guarantee. A/B ≈ 0.98 (n = 8); A/C = -1; A/D ≈ 0.96 (n = 7). Constant yields N/A. Spearman A/B = 1. Correlation does not establish causation.
H08: Classification. Actual rows / Predicted columns, No weight, First seen, Counts. Diagonal 2 + 2 + 2 gives 6/9 = 66.67% accuracy. Approved → Review is one record. For Row percentages use two decimal places to see 66.67%, 33.33%, 0%.

These are data files and instructions only. No application binaries, source code, personal settings, or production recordings are included.
