TsInversion

Description

The TsInversion enumeration is used to determine initial transformation method applied to variables (series).

It is used by the following properties and methods:

Available values

Value	Brief description
0	None. No transformation.
1	Log. Natural logarithm.
2	DLog. Difference of logarithms.
3	PCH. Increase rate.
4	Diff. Difference.
5	Normalization. Normalization.
6	Standardization. Standardization.
7	RateOfChange. Growth rate.
8	PCHA. The annual increase rate.
9	TS. Detrending.
10	SA. Seasonality correction.
11	OA. Outlier correction.

Comments

Log. Logarithm of the series points is taken: ln(X[t]).

DLog. The logarithm difference of the modeling variable nearest points for a specific period is found: (ln(x[t]/X[t-N]).

PCH. The variable value change is calculated in percentage to the previous period: (X[t]/X[t-N]-1)*100.

Diff. The series values increase is calculated: X[t]-X[t-N].

Normalization. The series points are normalized: x[t]/σ(x), where σ is root-mean-square deviation of the series.

Standardization. The series points are standardized: (X[t]-M(X))/σ(X), where σ is root-mean-square deviation of the series, M is mean value of the series.

RateOfChange. The growth rate of series values is calculated: (X[t]/X[t-N]).

PCHA. Increase rate for annual series values is calculated: (((X[t]/X[t-1])^S)-1)*100, where S is the number of periods per year.

TS. The trend correction is performed according to the formula: x[t]-f(x(t), S), where f(x(t), S) is series trend, S is trend type, determined by the ITsInversionInfo.Dependence property.

SA. The correction is performed with the Census1 method according to the formula: Iff(X>=0,SA Multiplicative, SA Additive), X are all series values. SA Additive(X[t])=X[t]-s1(X[t]), where s1 - is series seasonal component X[t]. SA Multiplicative(X[t])=X[t]/s2(X[t])*100, where s2 is series seasonal component X[t]. The seasonality type is also determined by the ITsInversionInfo.Seasonality property.

OA. The outliers are corrected by the K-sigma method: x[t]-f(x(t), k), where f(x(t), k) is series outlier x(t), k is parameter of calculation by K-sigma method, determined by the ITsInversionInfo.K property.