GEOG 3MB3 Lecture Notes - Lecture 5: Square Root, Normal Distribution, Multimodal Distribution
September 23, 2016
Distribution
- It is important to understand a group of data
- You may think this is just some kind of computational exercise but it is very
natural and is done by non-human animals as well
- The distributions of data can vary greatly
- There can uniform distributions, where no number is more likely to come up than
any other number
- There can be exponential distributions where there are great differences in the
data
The histogram
- Used to represent the frequency of continuous data
- It exists because it is hard to represent continuous data in a discreet form
- This number tallies the number of neighbours each country has
- This is not continuous data because these are all integers and discreet
- Histograms are not made for discreet data, so you interpret this histogram as a
bar graph
Birthweight histogram
- This is a better/more stereotypical histogram
- Seeing this graph we can intuit what this information means
- Graphing formalizes a process that we already do
Making a true histogram
- You start with a number of samples
- The green line is the arithmetic mean
- Each dot describes the difference between a number and a mean
- This helps you get a sense of the distribution
The Histogram (making one)
- Here you come up with a number of bins (basically the units/interval of change
on the x-axis)
- The first number is 7.4 so it goes between 6.75 and 7.5
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
It is important to understand a group of data. You may think this is just some kind of computational exercise but it is very natural and is done by non-human animals as well. The distributions of data can vary greatly. There can uniform distributions, where no number is more likely to come up than any other number. There can be exponential distributions where there are great differences in the data. Used to represent the frequency of continuous data. It exists because it is hard to represent continuous data in a discreet form. This number tallies the number of neighbours each country has. This is not continuous data because these are all integers and discreet. Histograms are not made for discreet data, so you interpret this histogram as a bar graph. Seeing this graph we can intuit what this information means. Graphing formalizes a process that we already do. You start with a number of samples.