◤Demonstration – TimeTransf◢
Main menu:
Computes the average value of the variable.
Computes the number of events in the series.
Computes how many time units the whole series lasts:
duration = MAX(date/time)/MIN(date/time).
Computes the same as duration (see above) but if you have defined a condition, it uses only events fulfilling that condition. If some of the first or last events do not fulfil the condition, conditioned duration will be shorter than duration. If events which do not fulfil the condition are only in the middle of the series, then duration = conditioned duration.
Computes the standard deviation. Standard deviation is the square root of variance (see below). Use this function when you need to know what is the average deviation from the average value.
Computes the minimum resp. the maximum value of the time series.
Computes the total value.
Computes the total of square values of the time series.
Computes the estimated values of linear regression b_{0}, b_{1} and the correlation coefficient in the linear regression function where Y is the estimated value of the linear regression function in time and b_{0}, b_{1} are regression parameters. Parameters are computed by the solution of normal equations using least square method where y_{i} are the particular values of the original time series and n is the number of events.
Parameter b_{0} is the value of the estimated linear regression function in time t = 0 (at the beginning of the time series). Parameter b_{1} is the smernice of the estimated linear regression function. Be careful when you interpret these values. The slope depends on the scale used with both values variable and date/time variable. Interpret this parameter b_{1} rather as: “the value of the time series increases/decreases on average by b_{1} in one time unit”.
The Correalation coefficient indicates the rate of correlation between the original time series and the estimated linear regression function. Correlation coefficient values range between <−1; 1>. Positive values indicate a positive functional dependence between the original time series and the estimated linear regression function, negative values indicate negative dependence. Values near zero indicate that there is functional independence.
Be careful while interpreting the results of correlation coefficient.
Compute a variance where y_{i} are the particular values of the original time series and n is the nuber of events.
Note that variance calculates with sqaure values. If you want to compare variance with the original time series values use deviation instead.
This function computes the frequency (period) spectrum of a series and then searches for the most occurent period.
Function Example measures correlation between a particular time series and the time series given as an example series. Example series is one of the examined time series. Enter the id number of the example series in the field next to the Function and Variable field in the Characteristics detail window. The result of this function is the correlation coefficient.
Correalation coefficient indicates the rate of correlation between the original time series and the estimated linear regression function. Correlation coefficient values range between <-1; 1>.
Function slope(5) resp. slope(20) searches for significant zlomy and returns as the result 5 resp. 20 major zlomu. To consider a zlom as significant, usek of data must be described minimum of 3 values. The algorithm scans sequentially the series for abrubt chage. This way it finds predem unidentified amount of zlomy. Then it orders the zlomy by the length of usek of the zlom and takes first 5 resp. 20 of them.
The result of this functions gives the number of siginificant zlomy, correlation coefficient between original time series and reconstructed curve from particular useku, slopes of the useky and finally the dates of events of the major zlomy.
Computes average value from all investment events. See functions invest count and investment. for details.
This function counts how many events did happend which we can oznacit as a investment event. If you want to know the value of three maximum investment events use function investment (see below).
Function Investment of a characteristics has rather economic application. This function constructs a typical month (31 day long period) and then searches for maximum of 3 events in the whole series with highest expense (a minus event). To see if this expense is significant enough we have to compare this expense with total expenses to accordant day in constructed typical month.
Typical month
Typical month function is constructed on 31 day basis. For each day the
function calculates the total of values accordant for this particular day of
month. This function also counts for each day how many events did happend so it
is easy to calculate average value for each day of month.
Payments function is analogically the same as Investment function (see above) but it works with positive values instead of negative.
Function day of payment returns date of events which has been considered as payments.
Function Wages of a characteristics has rather economic application. This function constructs a typical month (31 day long period). Then searches for a maximum of 3 days in the typical month with highest income (a plus event). See the Wages bound parameter on Parameters window. It indicates the minimum percentage amount of the total month income which must a wages have to be considered as a wages.
Function day of wages returns date of events which has been considered as wages.
Main menu:
Send comments about this site to the webmaster