Eli Katsiri

1
Lecture 3

• Eli Katsiri
• School of Computer Science

2
Excel Practice test

• February 8 (week 5) 7.30pm to 9pm
• February 15 (week 6) exam
• You will not be examined on the meaning of
  statistics. You do not need to interpret and
  discuss the results. You need to be able to
  produce the results using excel.
• Simple data fomatting and data manipulation
  exercises using basic mathematical,logical and
  statistical functions , autofilling, histograms,
  descriptive statistics and pivot tables.

3
Rounding (1)
71, 72, 73 and 74 are closer to 70 than 80
75, 76, 77, 78 and 79 are closer to 80 than 70
4
How do we decide?

• Decide which is the last digit to keep.
• Increase it by 1 if the next digit is 5 or more
  (this is called rounding up)
• Leave it the same if the next digit is 4 or less
  (this is called rounding down)
• Examples
• 3.044 rounded to hundredths is 3.04 (because the
  next digit, 4, is less than 5).
• 3.045 rounded to hundredths is 3.05 (because the
  next digit, 5, is 5 or more).
• 3.0447 rounded to hundredths is 3.04 (because the
  next digit, 4, is less than 5).

5
Negative numbers

• For negative numbers the absolute value is
  rounded.
• Examples
• -2.1349 rounded to hundredths is -2.13
• -2.1350 rounded to hundredths is -2.14

6
Rounding before the decimal point

• Is 123.5 closer to 120 or 130?
• Is 4278.3 closer to 4200 or 4300?

7
Rounding before the decimal point

• Decide which is the digit that we want to round.
• If this digit is bigger or equal to 5, then add 1
  to the digit on the left and make the current
  digit 0. Make all the digits to the right before
  the decimal point 0.
• If this digit is smaller than 5, subtract 1 from
  digit on the left and make this digit 0. Make all
  the digits to the right before the decimal point
  0.

8
Round MathTrig (3)

• Round(value,x)
• xRound(-1.582, 1)  returns -1.6y
  Round(3.1415, 9)  no changez round(123.5,
  -1)  returns 120
• round(4278.5,-2) gives 4300.

9
Trunc

• 5.6341432543653654
• 32.438191288
• 6.3444444444444
• To truncate these numbers to integers we
  consider only the left of the decimal point.
• 5
• 32
• 6

10
examples

• Trunc(2.1)8.4 returns 16.8
• Trunc(2.1)2
• 28.416.8
• Trunc(2.18.4) returns 17
• 2.18.417.64
• Trunc(17.64)17

11
Trunc

• truncation is the term for limiting the number of
  digits right of the decimal point, by discarding
  the least significant ones.

12
Mean

• The mean of 2,8 is 5
• The mean of 2,8,11 is 7
• The mean on 40,60,35,20,55,30,57,45 is

13
Round(mean)

• The rounded mean value of 40.60,35,20,55,30,57,45
  is 42

14
median

• 5 14 30 1 2 29 12 31 3 12 13
• 1 2 3 5 12 12 13 14 29 30 31
• M12
• 5 14 30 1 2 29 12 31 3 13
• 1 2 3 5 12 13 14 29 30 31
• M(1213)/212.5
• 50th percentile of the data
• The lowest number that is greater than 50 of the
  numbers in a list.

15
Mode

• The most frequently occurring number value in the
  sample
• If no data value occurs more than once, then the
  mode function returns NA
• 5 14 30 1 2 29 12 31 3 12 13
• Mode 12

16
Variance - Standard deviation

• 1,3,5

17
Percentile - percentrank

• The lowest number that is bigger than 90 of the
  data.
• PERCENTILE(data,k), k0.9
• PERCENTRANK( data, value)
• Returns the ranking of the observation relative
  to all values in the data set.
• E.g. p0.038 means that 3.8 of the data sample
  had lower values than our element.

18
Range

• The largest number in the data-set minus the
  smallest number

19
Comparing data sets in terms of their statistics

• Compare mean or median. The higher the median the
  better (higher) the data (e.g. higher the profit)
• Compare standard deviation . The higher sd the
  more variable the data.
• Compare histograms in terms of symmetricity and
  skewness. Do not explain what this means.

20
Logical