a \:b)@sdZddlZddlZddlmZddlm Z m Z ddl m Z ddl mZddlmZgdZd d Zd d ZdddZGdddeZdS)z+ Seasonal Decomposition by Moving Averages N)nanmean) PandasWrapper array_like)STL)convolution_filter)freq_to_period)rseasonal_decompose seasonal_meanDecomposeResultc Csptddt|D}|jddtddt|dddD}t|||}t|||}tjjtjt ||t ||f|||ddd\}}t d|tj|tj|j }|j dkr| }||d|<tjjtjt ||t ||f|||ddd\}}t |d|jdtj|tj|j }|j dkr\| }|||dd<|S) z Replace nan values on trend's end-points with least-squares extrapolated values with regression considering npoints closest defined points. css&|]\}}tt|s|VqdSNnpanyZisnan.0iZvalsr7lib/python3.9/site-packages/statsmodels/tsa/seasonal.py sz%_extrapolate_trend..rcss&|]\}}tt|s|VqdSr r rrrrrsN)Zrcond)next enumerateshapeminmaxr ZlinalgZlstsqZc_ZarangeZonesTndimsqueeze) trendZnpointsZfrontZbackZ front_lastZ back_firstknZextrarrr_extrapolate_trendsH  "   , r"cstfddtDS)z Return means for each period in x. period is an int that gives the number of periods per cycle. E.g., 12 for monthly. NaNs are ignored in the mean. cs"g|]}t|dddqS)NrZaxis) pd_nanmean)rrperiodxrr Az!seasonal_mean..)r arrayrange)r'r&rr%rr ;sr additiveTcCsT|}t|}|dur(tt|dddd}t|ddd}t|}tt|sVtd|drvt |d krvtd |dur|durt |}|}ntd |j d d|krtd d|d |j d d|dur|dd kr t dgdg|ddg|}nt d||}t|d} t||| } |dkrF|d}|d kr^t| |d} |drt|| } n|| } t| |} |dr| tj| d d} n| tj| d d8} t| j||djd|} |dr|| | }n| | }g}t| | ||fdD]"\}}||j||dqt|d |d|d|ddS)a Seasonal decomposition using moving averages. Parameters ---------- x : array_like Time series. If 2d, individual series are in columns. x must contain 2 complete cycles. model : {"additive", "multiplicative"}, optional Type of seasonal component. Abbreviations are accepted. filt : array_like, optional The filter coefficients for filtering out the seasonal component. The concrete moving average method used in filtering is determined by two_sided. period : int, optional Period of the series. Must be used if x is not a pandas object or if the index of x does not have a frequency. Overrides default periodicity of x if x is a pandas object with a timeseries index. two_sided : bool, optional The moving average method used in filtering. If True (default), a centered moving average is computed using the filt. If False, the filter coefficients are for past values only. extrapolate_trend : int or 'freq', optional If set to > 0, the trend resulting from the convolution is linear least-squares extrapolated on both ends (or the single one if two_sided is False) considering this many (+1) closest points. If set to 'freq', use `freq` closest points. Setting this parameter results in no NaN values in trend or resid components. Returns ------- DecomposeResult A object with seasonal, trend, and resid attributes. See Also -------- statsmodels.tsa.filters.bk_filter.bkfilter Baxter-King filter. statsmodels.tsa.filters.cf_filter.cffilter Christiano-Fitzgerald asymmetric, random walk filter. statsmodels.tsa.filters.hp_filter.hpfilter Hodrick-Prescott filter. statsmodels.tsa.filters.convolution_filter Linear filtering via convolution. statsmodels.tsa.seasonal.STL Season-Trend decomposition using LOESS. Notes ----- This is a naive decomposition. More sophisticated methods should be preferred. The additive model is Y[t] = T[t] + S[t] + e[t] The multiplicative model is Y[t] = T[t] * S[t] * e[t] The results are obtained by first estimating the trend by applying a convolution filter to the data. The trend is then removed from the series and the average of this de-trended series for each period is the returned seasonal component. NindexZ inferred_freqr')Zmaxdimz,This function does not handle missing valuesmrzJMultiplicative seasonality is not appropriate for zero and negative valueszxYou must specify a period or x must be a pandas object with a PeriodIndex or a DatetimeIndex with a freq not set to Nonez'x must have 2 complete cycles requires z observations. x only has z observation(s)g?rg?Zfreqr#)seasonalrresidN)columns)r0rr1observed)rgetattrrlenr allZisfinite ValueError startswithrrrr*repeatintrr"r ZmeanZtilerzipappendZwraprr )r'ZmodelZfiltr&Z two_sidedZextrapolate_trendZpfreqZpwnobsZnsidesrZ detrendedZperiod_averagesr0r1ZresultssnamerrrrDstE   &           rc@sleZdZdZdddZeddZeddZed d Zed d Z ed dZ eddZ dddZ dS)r a Results class for seasonal decompositions Parameters ---------- observed : array_like The data series that has been decomposed. seasonal : array_like The seasonal component of the data series. trend : array_like The trend component of the data series. resid : array_like The residual component of the data series. weights : array_like, optional The weights used to reduce outlier influence. NcCsR||_||_|durszDecomposeResult.nobsTFcCsddlm}ddlm}|}||r4|jdfgng} | |rJ|jdfgng7} | |rb|jdfgng7} | |rz|jdfgng7} | |r|jdfgng7} t |jt j t j fr|jj d} |jjd|jj| d f} nd|jj dd f} |t| d \} } tt| | D]\}\}\} }|dkr2|| n"|j| d d d |j| d dddt| d|}|dkrr|}|dkr|r|jn|j}|||| q | | S)a@ Plot estimated components Parameters ---------- observed : bool Include the observed series in the plot seasonal : bool Include the seasonal component in the plot trend : bool Include the trend component in the plot resid : bool Include the residual in the plot weights : bool Include the weights in the plot (if any) Returns ------- matplotlib.figure.Figure The figure instance that containing the plot. r)register_matplotlib_converters) _import_mplZObservedrr0ZresidualrAroZnone)ZmarkerZ linestyle)rrz#000000)ZcolorZzorderr@)Zpandas.plottingrMZstatsmodels.graphics.utilsrNrIrr0r1rArDrEZ DataFramerFrr-Zsubplotsr6rr<plotr5 capitalizeZ set_titleZ set_ylabelZset_xlimZ tight_layout)rJr4r0rr1rArMrNZpltZseriesr>ZxlimZfigZaxsrZaxZdef_namer@titlerrrrQs6       zDecomposeResult.plot)N)TTTTF) __name__ __module__ __qualname____doc__rKpropertyr4r0rr1rAr>rQrrrrr s(       r )r,NNTr)rWZnumpyr ZpandasrEZpandas.core.nanopsrr$Zstatsmodels.tools.validationrrZstatsmodels.tsa._stlrZ#statsmodels.tsa.filters.filtertoolsrZstatsmodels.tsa.tsatoolsr__all__r"r robjectr rrrrs$    +