Time Series for Android

trattare campioni di variabili aleatorie di tipo serie storiche

The application consists components processing samples of random variables of type time series. Applied five methods for prognosis and calculate the autocorrelation for disclosure of internal cycles. This application stores the data of only one sample. Enhanced application capable of storing data samples and many more features of this app is TimeSeriesResercher.

Processing includes storage (edited, deleted, renamed) a sample of a random variable, calculates and stores the basic statistical characteristics: sample size; minimum and maximum velues; arithmetic mean of sample; mode of the sample; median of the sample; lower quartile of the sample;upper quartile of the sample; lower decil of the sample; upper decile of the sample; average absolute deviation of the sample; standard deviation of the sample - biased estimator; sample standard deviation - unbiased estimator; variance - biased estimator; sample variance - unbiased estimator; coefficient variance of the sample; range of variance of the sample; interquartile range of the sample; inter decile range of the sample; quartil deviation of the sample; decil deviation of the sample; sample skewness; sample kurtosis; check zero skewness is rejected or not rejected(small sample); check zero kurtosis is rejected or not rejected(small sample); normal distribution rejected or not rejected(small sample); check m-test:zero skewness rejected or not rejected(for large sample); check m-test:zero kurtosis rejected or not rejected(for large sample); check m-test: normal distribution rejected or not rejected(for large sample).

The application along with the calculation of the aforementioned basic statistical characteristics has a check for normal and uniform distributions according criterion of Pearson, calculation of limits, such that the percentage of the population with a given probability falls to them. This application has functions to calculate (visualization and storage) autocorrelation and five type predictions of samples of type time series. The application has a help - the formulas for calculating the above-mentioned statistical characteristics.

A sample, results from its processing and forecasts can be saved in database (Sqlit). Tables with these data can be exported for printing for example, by means of Sqlit browser. The database may be exported in /mnt/sdcard/timeseries.db (or in the Phone storage) and back import. So the database can be stored in other media.