EDA - Visualisation

EDA - Visualisation

1. Basic Plots

2. When combining categories in histograms

   • 1. Combine the 2> values
     • 2. Divide the values by the number of bars being represented for that range
       • 3. This will create an averaged set of bars across the range (Changing the shape of the histogram)
   • Bars don't have to be the same width - Its all about AREA.... Not HEIGHT

3. Distribution Shapes

   • Distribution Types
   • Bulk of Density
   • Modal Values
   • irregular

4. Relative Frequency Density - Asymmetrical

5. Outliers

6. Measures of location

   • Mean
     • Add all the values and divide by the number of values
   • Mode
     • Most common value
     • Modal Class vs. Mode Tables. Can appear differently. Class is important. The resolution/number of bars can better determine the mode
   • Median
     • The middle number
   • Variance = average of squared differences about the mean
     • Always divide by one less than the sample size - Don't include your mean
     • Measures the spread of the data
     • Variance Square Rooted = Standard Deviation
       • How spread out the data is
         • Zero = Max & Min's the same
   • Skew - 0 = Symeteric
   • Kurtosis
     • Tail Fatness - Higher numbers thinner tails

Imagens e arquivos do mapa