Stats

Document Sample
Stats Powered By Docstoc
					x            y
         1           10
         2           20
         3           40
         4           80

new x        new y
         5
         6
         7
         8
         9
        10
        11
        12
        13
        14
        15
                                100
                                125
                                200
                                175
                                200
                               8888
                             797979
                                123
average                    100973.8
average if less than 300   153.8333




                              10000   Central
                              12000   Central
                               1100   South
                               1250   South
                                300   North
                                200   North
                                220   North
average                    3581.429
north average                   240
central average               11000
south average                  1175
               rank        rank
values         desc        asc
           1           9           2
           3           8           3
           5           7           4
           6           6           5
           7           5           6
           8           4           7
           9           3           8
          14           2           9
         100           1          10
          -5          10           1
         Q1         Q2         Q3
growth        10%        60%        20%   Average growth          30%
                                          geomean growth     0.228943
                                          geomean manually     1.200% numbers multiplied together
                                                             0.228943 number raised to power of 1/(nu
numbers multiplied together
number raised to power of 1/(number of elements)
           group1    group2    group3
                 100        80         3
                 100        94        45   1.990071 peaky!
                 100       100       106   0.224025 a little peak
                 100       105       120   -1.11187 flatter
                 101        85         5
                 101       107        19
                 101        98        90
                 101       102       119
                 102        90        20
                 102       100        95
                 102       100       150
                 102       109       190
                 103        92        25
                 103       102       100
                 103       103       180
                 103       110       210
                 105       120         5
                 107       121         7

std dev.    1.847096 10.63844 69.60714
array
          1   n largest    n
          2          234       2
          4
         56
         78
         99
        234
        543
 120
 123
 124
 124
 155   254.4444 average   total/number of values
 166        124 mode      most repeated
 177        155 median    middle value - average of two middle numbers if even n
 180
1121
two middle numbers if even number of values
array
         1
         2
         3
         4
         5   minimum value                           1
         6   first quartile    25th percentile    5.75
         7   second quartile   50th percentile    10.5
         8   third quartile    75th percentild   15.25
         9   maximum value                          20
        10
        11
        12
        13
        14
        15
        16
        17
        18
        19
        20
group1        group2
          1            1
          1            2
          1            3               1.491324 tail extending towards positive values
          2            4               -0.81898 tail extending towards negative values
          2            5
          2            6
          2            7
          3            8
          4            9
          5            9
          6            9
          7            9
          8            9
          9            9
         15            9


 3.925497 2.898275 stdev - how widely scattered values are from the mean
 4.533333       6.6 average
       14         8 range
  3.79239       2.8 stdevp - just like stdev but when you are calculating for the entire population - not just a sa
 15.40952       8.4 variance - variance of a sample set
 14.38222      7.84 variance of population - not just a sample


                           variance - mean of squared deviation from expected value (mean)
he entire population - not just a sample
                      product
                      accord     civic     odyssey pilot     CR-Z        Crosstour Insight
JM Family Level   A          100       110       150     200         5         200          5
                  B          105       115       155     205        10         205         10
                  C          110       120       160     210        15         210         15
                  D          115       125       165     215        20         215         20
                  E          120       130       170     220        25         220         25
                  F          125       135       175     225        30         225         30
                  G          130       140       180     230        35         230         35

                                         relative position
                      Level    E                   5                     Bonus
                      Model    odyssey             3                             170


                                         Match and Index example
Ridgeline
       160
       165
       170
       175
       180
       185
       190


             offset
                      215
In other words, recent observations are given relatively more weight in forecasting than the older observations .
Newer data is weighted heavier!
damping factor close to 0 has more smoothing (prettier, but less "True")
ng than the older observations .
multiple R
             http://webhome.broward.edu/~mfenick/Cabinet/stats/coreye.html
asset return (random variable)
dispersion of values is variance: useful for one variable

two random variables: strength of relationship, covariance.

variance and covariance are convenient but not intuitive

variance is turned into std (by taking square root)

covariance - translate into correlation coefficient runs from -1 to 1.0

unitless - closer to 1, greater strength of relationship of random variables.
array1                      array2               matrix
         2              4            5       1            18           14
         1              5            2       3            15           16

a            b              e            f       ae+bg         af+bh
c            d              g            h       ce_dg         cf+dh




matrix multiplication

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:7
posted:10/26/2011
language:English
pages:19