Day3_​S1_​Irregular_​Sampling

1 · Time Alignment

Read Day 2's table and sort each pump's readings by time.

Step 1 · Read and put in order

1. Read the Day 2 monitor table with CSV Reader
2. Turn the text timestamp into a real date-time with String to Date&Time
3. Sort by pump then time with Sorter

2 · Regular Sampling

Collapse to one row per pump-hour, then fill the gaps.

Step 2 · Make the sampling regular

1. One row per pump-hour, averaging duplicate readings, with GroupBy
2. Fill the hours that are still empty with Missing Value, linear interpolation

3 · Smoothing Per Pump

Loop per pump so a window never crosses machines.

Step 3 · Smooth inside each pump

1. Start one pass per pump with Group Loop Start
2. Take a 24-hour rolling mean with Moving Aggregator
3. Collect every pump back into one table with Loop End

4 · Quality Gate, Explore

Score quality, then profile columns and coverage.

Step 4 · Quality gate and profile

1. Score quality with Rule Engine, split with Row Splitter
2. Profile columns with Statistics
3. Per-pump coverage with GroupBy, one colour per pump with Color Manager

Explore before you decide

Profile the aligned signal before Step 4 locks any threshold.

Explore · Profile before locking

1. Profile columns with Statistics
2. Per-pump coverage with GroupBy, one colour per pump with Color Manager
3. bearing_temp_c: spread with Histogram, over time with Line Plot
4. HIGH vs LOW with Box Plot, against flow_m3h with Scatter Plot