Types of data

The different columns in a features matrix can be any of a number of different data types, including the following:

• a numerical value, e.g.:
  • the area of a leaf
  • a temperature recording
  • a whole number (or integer), possibly constrained within some range, eg:
    • a count of the number of leaves on a plant
    • the grayscale intensity of a specific image pixel (0-255 for an 8-bit image)
• simple text data or strings, e.g. a name of a person, place or thing
• a categorical value (can be recorded as a string or an integer), e.g.:
  • a plant species and/or genotype
  • an experiment number
  • a healthy or diseased label
• a date or a time – often referred to as a datetime – this could be:
  • a fixed point in time, often called a time stamp (e.g. 11:06 on 19th July 2022)
  • a fixed specific period of time such as a year or a month or a day, often called a time interval (e.g. July, or the year 2022, or the day of 19th July)
  • a measured amount of time, often called a time delta (e.g. 13 days, or 14.6 hours, or 285 minutes). For example, number of days since sowing.
• a true or false value or yes / no value, a.k.a a Boolean. E.g. was some experimental treatment applied or not?

Numerical data can be whole numbers which are referred to as integers, or numbers with a decimal point, referred to as floating point numbers or floats.