Comet wants a quick answer. "Which plant drinks the most water? Beacon has kept a log for a whole week."
She sorts the log by plant name and frowns. There are entries for Tomato, tomato, tom. and TOMATO.
"Beacon thinks we grow four kinds of tomato," she says.
Wren scrolls further. "What do you notice? Some rows have no water reading at all. And this one says the soil was 140 percent wet."
Nova circles the screen. "Data has to be cleaned before it can teach you anything," she says.
Comet sighs and opens a fresh page. "Fine. Clean first, answer second."
Information is the facts and patterns we pull out of data. Raw data often needs work first.
Data sets can need cleaning. They can be incomplete, with missing values, or invalid, with values that make no sense.
Cleaning data makes it uniform without changing its meaning. For example, every spelling of tomato becomes the same word.
| Row | Plant | Water reading (0 to 100) |
|---|---|---|
| 1 | Tomato | 62 |
| 2 | tom. | 58 |
| 3 | Basil | (blank) |
| 4 | TOMATO | 140 |
| 5 | basil | 35 |
Beacon water readings run from 0 to 100. This log is invented for our story.
Metadata are data about data. For a photo, metadata could include the date it was made or its file size.
Metadata help people find, organize and manage information.
Changing or deleting metadata does not change the primary data. Fixing a wrong date on a photo leaves the picture itself alone.
| Statement | True or false? |
|---|---|
| The date a photo was taken is metadata. | ? |
| Deleting a photo date also changes the pixels in the photo. | ? |
| Cleaning data should change what the data means. | ? |
| A missing value makes a data set incomplete. | ? |
Wren notices a pattern. On days when the crew took more photos, the tomatoes were taller.
That is a correlation: two things change together. A correlation does not prove that one causes the other.
Maybe the crew takes more photos when the plants already look great. More research is needed to know.
One source is also often not enough. You may need to combine data from several sources to reach a conclusion.
People choose how data is organized and where it is stored. Those choices affect speed, reliability, accessibility, privacy and integrity.
Beacon could keep every reading in the outpost, or send everything to mission control. Each choice trades one strength for another.
Tidy work. Tomorrow in the Beacon Lab, you will compress a greenhouse photo by hand.