## Chapter R software helps convert data into the

Chapter 1 There are so many data sets available which involve time components and these type of prediction were neglected often because of the time formats and made it very difficult to handle. Thus time series forecasting became important area of machine learning. A time series is a set of chronologically arranged data points collected over time. The data points are recorded in the order of their occurrence.

Time series analysis is very important in forecasting the future behavior and analyzing the past behavior of the datasets and comparing datasets with other time series as well. It is particularly important while dealing with yearly Government budgets, monthly sales and advertising expenditures, and temperature distributions. Most likely, you will be working with data that already exists or the data that has been imported from other sources. Often, such data is not in your preferred class or format. You will be struggling with data which is not in the frame that you need.

We Will Write a Custom Essay Specifically
For You For Only \$13.90/page!

order now

The R software helps convert data into the preferred type of data. It provides you several packages to support your forecasting analysis and build forecasting models in time series such as ARIMA, GARCH, and EGARCH. This chapter discusses the general time series data and the other types of data that are converted to time series data, followed by the time series models. In this chapter, you will learn to Use time series data Convert any data to time series data in R Forecast time series models in R Time series are used in many domains including weather forecasting, finance, and earthquake prediction.

The datasets and the models that were discussed before in the Machine Learning with R – Student Guide arise in the situations where observations are independent of each other but in case of time series data sets observations are dependent. Here we have to deal with the change in frequencies with respect to time. The given example of time.series data explains a time series in which the observations are dependent on the previously observed data. In this example, it is necessary to develop time series models. Since the time series is sequential in nature, the time stamp is displayed in years and the average speed in km/hour. time.

csv, header FALSE) Average_speed – ts(time.seriesV5, start 2010, frequency 12) class(Average_speed) plot.ts(Average_speed) Figure 1 Average speed with respect to time The time stamp may be displayed in nanoseconds, seconds, minutes, hours, days, or months. Any case of dependent variable or a variable of interest which we observe on regular basis is treated as time series. For example, the data related to economy, businesses, GDP, and unemployment, which is used to make future predictions, is stored in the form of time series.

Thus, time series plays a crucial role in various domains. An observed time series can be decomposed into systematic and non-systematic components. Systematic components These components of time series are consistent and can be described and modeled directly. Trend (Secular trend or Long term variation)It is a long pattern that produces irregular effects. It can be positive, negative, linear, or nonlinear. It represents the variations of low frequency. The high and medium frequency of data is filtered out from the time series.

For example, if we have 10 years data of weather reports for a particular country, then long term trend appears and if we have data for almost 100 years of weather reports for the same country, then very long term variation will appear. Cyclic (Cyclic variation) It is a pattern that follows a specific trend of up and down movement. The duration of cyclic variations depends on business problems which needs to be analyzed using time series.

Seasonal (Seasonal variation) It is a pattern that reflects regular fluctuations which are related to periodic cycles. Many time series data indicate seasonal fluctuations in monthly and quarterly cycles of the year. Non-systematic components These components of time series cannot be modeled directly. Residuals (irregular variation) These are the residuals of the original time series after the trend and cyclic variations are removed.

They are unpredictable like accidents and earthquakes. They are also referred as noise or errors which can be removed by time series techniques. There are some techniques available for stationary time series as well which assume the mean and variance to be static as required by future prediction analysis.

If the time series is non-stationary, it is first converted into stationary time series and then further statistical analysis is done. R programming software facilitate us with many default functions and packages, which convert can convert any type of data into time series data. We store that data into R is defined as time series object.

We can use timeseries object functions to create, modify, and analyze the behavior of a time series. Function ts() creates the time series objects in R. The basic syntax for the ts() function in time series analysis is Timeseries.data- ts(data, start, end, frequency) Here, data is a matrix or vector containing the values used in time series. start indicates the start time for the first observation in time series. end specifies the end time for the last observation in time series. frequency specifies the number of observations per unit time. Consider an example of a vector representing the temperature for plotting a time series graph.

Temperature – c(15,34,56,34,67,34.5,98,56,78,23,10,23) Temperature Convert this data set into time series object using following code. temp.timeseries – ts(Temperature,start c(2010,1),frequency 12) temp.timeseries This is the output produced by the code. JanFebMarAprMayJunJulyAugSepOctNovDec2010 15.

034.056.034.067.034.598.

056.078.023.010.023.0 Next, plot this data by using the following code. plot(temp.timeseries) Figure 2 Time series plot using ts() function You can also plot multiple datasets of time series in one plot by combining multiple series into a matrix.

Consider two datasets in the form of vector in R for creating multiple time series plot. Vector1 – c(15,34,56,34,67,34.5,98,56,78,23,10,23) Vector1 Vector2 – c(25,56,78,65,90,100,132,200,213,78,65,34) Vector2 Next, convert the datasets into to a matrix. combined.vector – matrix(c(vector1, vector2),nrow 12) Next, convert this data set into a time series object. vector.timeseries – ts(combined.

vector,start c(2012,1),frequency 12) vector.timeseries The code produces the following output. Table 1 Table showing multiple time series objects Next, plot this data by using the following code. plot(vector.timeseries, main Multiple Time Series) This code will generate the following plot. Figure 3 Multiple time series plot using ts() function Frequency parameter in ts() function defines the number of observations per unit time or the time interval between the data points.

Frequency 4, specifies the data at quarterly basis Frequency 5, specifies the data at daily basis (business days) Frequency 6, specifies the data at every 10 minutes of an hour Frequency 12, specifies the data at monthly basis The only constant parameter in time series is data. There are many types of data are collected over the time like quality measurements, stock prices, weather reports, population growth etc. For this time based data needs to be forecasted by time series models.

Time series use those models for forecasting which are based on previously occupied values in your data. This section discusses about the given topics time series functions Construction of time series objects Application of time series models The zoo package provides a set of powerful tools to manage ordered observations. It aims at performing calculations containing irregular time series of numeric matrices and factors. Zoo also supports weekly regular series which have an underlying regularity.

The time series object ts() has certain limitations with unequally spaced data. It cant be used to represent the daily level stock prices because the stock prices are equally spaced for Monday to Friday, but not for Friday to Monday or in case of market holidays on weekdays. On the other hand, zoo is flexible and performs easily on unequally spaced data and numerically indexed data.

You can install and load the zoo library by executing the following code. install.packages(zoo) library(zoo) Let convert the given list into date format using as.Date() A.Date – as.Date(paste(2003, 02, c(1, 3, 7, 9, 14), sep -)) Note as.

Date()converts character data to calendar dates. The format isas.Date(x,format), wherexrepresents the character data andformatrepresents the desired format. Univariate plotting using zoo Univariate time series plot is the simplest form of plot. It represent the single line of observation recorded over regular observations in a single plot such as monthly returns etc. Here we are creating a univariate plot using zoo()function for randomly distributed numbers and a vector which we have created above i.e. A.

Date. A – zoo(rnorm(5), A.Date) A2 – zoo(rnorm(5, sd 0.2), A.

Date) plot(A) lines(A2, col 2) Figure 4 Univariate plotting using zoo Multivariate plotting using zoo Multivariate time series plot is a form of plot with single panel having first line of plot as the main curve line and remaining curves as functional lines. It represent the multiple lines in a single plot. Here we are creating a multivariate plot using zoo()function for randomly distributed numbers and a vector which we have created above i.e. A.Date.

z – cbind(A, A2, zoo(rnorm(5, sd 0.5), A.Date)) colnames(z) – LETTERS13 plot(z, plot.type single, col list(B 2)) plot(z, type b, pch 13, col 13) plot(z, type b, pch list(A 15, B 3), col list(C 4, 2)) plot(z, type b, screen c(1,2,1), col 13) Figure 5 Multivariate plotting using zoo Barplot using zoo We can also create a bar plot using zoo object.

Bar plot basically represent the numerical data with rectangular bars proportional to the value they belong. zoo() function. A – zoo(cbind(rpois(5, 2), rpois(5, 3)), A.Date) barplot(A, beside TRUE) Figure 6 Barplot using zoo xts is also an another type of time series object which provides uniform handling of time based data by extending zoo.

It boost the endemic format of information and allows user level custom make changes and extension. It carries all the features of a zoo object. It consists of a matrix and index which has to be time-based. The xts objects can be thought of as a matrix of observations combined with an index of corresponding dates and times.

Construction of an xts object install.packages(xts) library(xts) Create the object data using 5 random numbers. data – rnorm(5) Create dates as a Date class object starting from 2017-01-01. dates – seq(as.Date(2017-01-01), length 5, by days) Use xts() to create data1. data1 – xts(x data, order.by dates) Create data2 (1899-05-08) using a POSIXct date class object.

data2 – as.POSIXct(1899-05-08) Create data3 and add a new attribute called born. data3 – xts(x data, order.by dates, born data2) Output Table 2 Output table of xts object In the above example, we have created an xts object or a time series object for further analysis.

This xts objects results as a matrix with index of corresponding dates and times. Note as.POSIXct() function convert an object data into date or time format. The first step in time series analysis is to extract trend, seasonal, and cyclical components. Linear filters is one of the methods that is used to apply filters to extract trend from the time series. The first step in time series analysis is to extract trend, seasonal, and cyclical components. Linear filters is one of the methods that is used to apply filters to extract trend from the time series.

The function used for finding filters is filter(x,filter) where x is the time series data and filter is the coefficient required. Lets take an example time_series data in time series and plot the weekly and monthly averages. The function used for finding filters is filter(x,filter) where x is the time series data and filter is the coefficient required to find the average speed. Now we will take an example of one of the variable of time_series data set, we will convert variable V8 i.e. AvgMovingSpeed of time_series data set into time series variable and name it Average_speed. Then we will plot the weekly and monthly averages.

time.series – read.csv(C/Users/Amita.9.Jain/Desktop/DA/BOOK/time series code/time_series.csv, header FALSE) time.series Average_speed – ts(time.seriesV8, start 2010, frequency 12) plot(Average_speed, typel) Average_speed.

weekly – filter(Average_speed,filterrep(1/2,2)) Average_speed.monthly – filter(Average_speed,filterrep(1/10,10)) lines(Average_speed.weekly,colpurple) lines(Average_speed.monthly,colgreen) Figure 7 Average speed of time_series data using linear filter (NIIT Time series modeling 1.

PAGE MERGEFORMAT 10 (NIIT Overview of SQL Server 1. PAGE MERGEFORMAT 3 (NIIT Time series modeling 1. PAGE MERGEFORMAT 11 Objectives Exploring Time Series Note General Time Series Conversion of Any Data to Time Series Data in R Multiple Time Series Different Time Intervals Frequency Note Forecasting Time Series Models in R Constructing zoo Objects Constructing xts objects Linear Filters [email protected] /3K/IS3/_7wYm xb1ILf3bXQYvGp( 6DW7w(l,P_8vk66,b-6)(_7ctNNIQN//feu G4vVNVXvuJuuphj89yIcXcotE24j5loqZutvvvEv z [email protected]_o _1Xc8ZyhV [email protected]@DqpAM [email protected] [email protected]@@ BZ1x @([email protected]([email protected] W @(@BpE(2n6 sB6w2HBWAEBgE(W @(@hAFPA zh KhAq 3hNG(@(@hH(BZE tMeWy o.MPPdDBqBPWB / oxJP6BNy DWgEBxYOWII8/7R ,DvDqDd,[email protected])PGdoBi7I8gbADrXjV4 PaY,xs__xJ4cE(zdzPBRMSH/Cdi3RRp2 @[email protected]_gL76K2FVxq, 49rcW [email protected](JBywGDsgfJy, [email protected] oAilO OELZYaP1 BOT FzaQ,GgVH [email protected]@ZGCx P(_. dijqpuH8d)y44kat([email protected]/d )Ivc4bFk 7nkFvq7D(H/[email protected]@h6,[email protected] [email protected] PCpGC- ,ad)CD af I iPvpY-P.

[email protected]_CqAR cHEo Y40 )@[email protected](r H3Aj @ [email protected],6E6G z,nR d [email protected] JxVsBWd)dOJ 43si 2xbFd PB [email protected] bnGZJ3XyZ m typ5HvCBT509X.,orE8EOJG [email protected] [email protected]()kGG fDiXiAy(uH37D_q/J/I/WjO7J1_A8(V2e- upsW(m6wdDCgCQ .a_,.)GHEsp.

P_V4,[email protected]@[email protected]@ 088BQhCTQCXMp286( /JKNlGxfRCiw sO_yIX( ghgc.Dv6MXhmwaJ0QL(O PeTP.

[email protected]@u-h)XuR.J/[email protected] TP)[email protected](vEvQg4r,[email protected]_yp_M DDUt9M4PAikXN9 f 9oGrm-/[email protected])WG0Z.r Tpb5)kcs3,AwQvKz A(wEDg.08 t-rL bS.(A X )FDDU4uQFH Z8V j(ZqKhe/jnLY)fQjDrAnqv398B1N (U0R1fNNzSX4b-1SucG23PidLz T6/wAX71VBWi,S) QC5/Om ,sPJeKaC45K( RAJfgYafmYzdDGd(W0eGO5yE1v_EH ThpjPAoV Fj.(/zv [email protected]@Bm([email protected](U8NFu – )XsvB4Kb5R)FBLS))5cP1D [email protected] [email protected](JywD31VkTaRPhY BuAe(-PE _)R9R6UjCbuSquGAV,eR-NXg_88 [email protected] @WouA(B PE)cFc/7ZH/mORVfVmaxbwbTMvPsH5C5kGQ) nB/Vr48obQga R7MW)[email protected])cPSA7ft nCnPlxeZEtEi_nd3nt7m 5KDTT4bROxyc1tRm-3GL0,[email protected]@[email protected] 3R4jPAF YSE(b4_.QMfmf6x Sc5A.

.UxlXHjFWJog/2j3-ib9R,[email protected],_5(Z03l-iLmdWd_nDq g NWuoVonL(rf/USJ)ZNQP4Vk(BE/DIoK- )lH9IwX1t 3223t, [email protected](oz1qtFrDGjcVo-p-f_Vo,SNA/ 3cLxbr6R2EG.

_ywuvTJZ8Art [email protected],)TVrpoYsMMx u_j5nYNGcg9YMslf7exDuMi55wDg.3rvx uz.NiDN xvm l-usUD SSh3BBFZS(tTdrPJBeuEfJE-MtEMOu7f0d_/pt6Yyrd K5-RG)W2.5O/[email protected] (( zfwO IVm hPBRGJOFeYK/CYaznfF6aozrslROVWVQPTZXJ ldxWe ZMV2jxDo9xrMs2KmIkmDo98KkD.6 /ZQ)x,cC iBr7c,(L z E6fQMmhoWP6,.JXNs) IrZnMiyHu(iG3 zfoYAp bgu69gh q0Ah(zT8xAyQBBk) OCQQ u7IPcek_wK5kwad, MH(yf0M VE)W0y6jAyAH.3EzEn-aQBsM 0bv .

[email protected],Nrdh-eGUwF2 [email protected] hu2S6(hElB P 0FU )OwI2_zCa N1Jg5tsldBEy0hi.MXRe,f/e ET4MzZwO4A LhB DB_1D7xyJmYOcZ,Wzdt vmZFXi,[email protected] [email protected]@H)0M26QZDae(P 2ym.XQu(.

3V(n hFQR/jva2 H92)G Nh5bcbbcQDUR8u6YZUQBoj D0JQyBBD)[email protected])arWHo9()B,AC UBPUmCCSsBS3G39 (4e8r5Deb 1MQi gE_20 cL7/L2Xrg14SDn B5cf81pr/0lMurvs_Cck6 T 1e9j)Htft4OIHGCcRK2uC0 yiA8t iFNuSICd.e9.)[email protected])AvuOWx).M,tRW9BF h1QZC [email protected],8B X42vxZ -bcHJlPD97L4tGLJA Kq(,YPJ1RwQDQl uub Mb6 NWc)CfdimaWNAo V L8XVbzQ D_RBKW G (,yOZw2, 0A7)PaPuBVWV OYjSX) Cr ROK759wUA7wN(Y5Yc_,d5SUu0SBSvO .

Y/)Cl-qW0O-jCe2fKTXPl [email protected]@LOz5nBb15wO(mXBtH/G dWeSP PaJgJFgE1YJsRTrvuVwXOPWneR-hfc0CG42([email protected] a/,t,2O)QLVjnJA.6rC3(Wm,DbUftEtMpuXQ E5e_B9f1J(@ 0xp9O(3 b_5[email protected],aost4RihB 6HMSNnLKYlTb-pLf z K3FV7F8pyJEA)9 ,ea7u3Jm 9HL/J5,RrE3emZ5uTApLzzy_67_wjX9GlxXU VVFVdBmqX8zFx0,bIgm6g XCHE)BOs sn3Ud7fyc3YZW. ,70a2bOsQgbBb5x [email protected] 7 D(XzBE)v0A_ eqH-sA9sy-RYCv6EDw xuCsdxOzLxsX.

[email protected],ky_ev 8Pr8 qDkFwm eAFuag8vwbq54ze/71QxpE.3lX0b5Gm5z1,n8wqhw021iOon,W90J/[email protected]–rBv4)o 2KjZ.Z/5SnBWliyelhUjfclTgYyIYOHLJ J NNIILIENwZnPF4U,yp _.ohk 5pxKAteBBxJH C24XQ KkUgH.toVnvOw-lkAGSZjDuUZHZCYx/tOkefl9f,.

(IcHWXAKOQ1Yg,58048,(o,ZjDZkFRHWY3peHRYo9SsfgTyepeoi9G_Uyh45t(uD OnUetNy3WqZ19pYUwJEHE(yhLikV8oFbkF.ldnYM,UOGXld7enckvgwqGxdc5YN3eIMu(O3n7jk Gny2HC6Q_HgNOH3twBJwY.s-wn,9I_(3BKEauD)AwA)1o9-FO5q5oF_LeB(BX(@(IyRyQ0 VlUiVrgyq7 Tq2m___pFwut7fNAoYvZ_999f8nu)vLYU0l_WbAecrKFg,bdFIbxR3_gWX6hiFr9gr/1U9(ZDj8wMeex0vP([email protected]_u SaYfnO_o5vQk_y [email protected]@kPJB FUQF)AP1j)4BmSSrsJCJV-q -/j.3 @@@hG((tZbU5p5BOA5RsB3e_6uX77OuSQz0l-lre F891mRRyZDhNEdazadMRBxqIC3/g. Hv @3l)QZ1J e EOd DA1aKqSZ Tc)t(6(8RdWDxCm.XeF4Fb9a_E_4kh B/h0(a t8y2 tQ rz4bvP6,Y,)JjO 0e5SmxPfaxti .

HBHJvS0dgu/OuAZUH HT g 4PDq_uWlh / _u4w [email protected],[email protected][email protected])B /._UGteD2BrBky/Zzj3,Bw.U-6VjTGOs cp4MB8zePBQABwjVJtaByMBpB 9 [email protected],JN 5qU-DX(DgtvMhCM6Nud(22Pfeg.

l7l2TH ePG52zVPGBgvcrMRG7xVtV2p 9y5S 7a([email protected] eFEw8)7tn2Wysvf)VSLcmfPJh0hdPp,[email protected]@3-d)[email protected],Ft uABCQjcBxGY3jCgEpXIJ3VzugtvntA8vz,,lttg-mDNBBB8MMEB H(@H(8dmuM)[email protected] xtSld1ufGOQs6nFimamyam nR [email protected] (4 dq bZn)[email protected]@Z0V5PW)pfe1-_Li/ MoeVXkSV.-,7Xnv0EAhd T5t5P(vWZRYQjDa8EH([email protected]/fM4/19qLR3i0nc .JY_sFQ7_B8 wB z7OcqTdvNnpZi8l.SikAd2(Mq4AYw3fI(4 U,sYMZyTGNO(3yGSe2Ai )PDPI.ynB) i/ RsQryPgwQTGQBBj43AMzj.

E(CV4s)l)[email protected])p-H6F7JTe9XfHP [email protected](gEB 1FeOeSfNk dBPU [email protected] EIZN 67St tw0sOBwWVdBC([email protected]))JE,usQBtX)DIPG CBqXX5z1ynY3s wG8s _v1gQAu Yz(R8H z(7.KQ)RPi [email protected] o Rszol-l,0cBYmznW_WQXR_D-5GkhEAuLuy1AhB qdwbVJQSTXBJj sQ 7 4wt7ta( , m7,MTl5vIQ5I2N- EJP ThA zOVx [email protected], V4d8pHIQe [email protected](kAzLx9ddJFqK8E [email protected]/[email protected]@rfcPBGOd4eIQDIQByPT )(D 4 57t5mBv /pX GR_LxptsAN6bpa)BIh xd [email protected])ypPQAQjtP(azTrD5RlDs9-YrU 9E)EQH,DYDKv)[email protected](@ f-64t2fX8BhsGv1w3rQbOjzBx(BBSY75yQA 43jE kHlF [email protected] MpXrEkQXy6m-.VBGVPvsUrJ d 9STAK2bfpX,DHykuXy_QqA txidPul.(HE1h GT82AsTdVb(J) [email protected]@dFgJ pz [email protected],ETB0xbL_UsHK/@9(xFBBT/99AA81f7Q-j RQMA-Ev qbon m UYY(iF KPGa @HY5.

Aj .jA(bBfDhNYgmdp.z qqPPzAawR.5zD6VWTBy4GxGn(P13BV0 )U(iAV/n5Byp R(hAWX) [email protected] _48q4HS0tNQ2P/ABqFAB-kQLa B6erz,SB TL.B6_1eIy2yBAae /z1RZp KSDDDUFpxSHBOR56mWV t25hgs0 mkhBHMQbG4 EiWkxs0ReBBDZ)Ti5VpLkjtY.akFd 8Q.DpSbpkYXh(SeFM-/5ObXiA((UjKA GB,Gy)DP-fQ A 4(q3.UHD2([email protected]

[email protected] 42R5L iA(dQ 5 9QOwX)[email protected]@)xujk)zcAjTQAz4BM nNqDq p @[email protected] J4lnDB BtXH4M4)4dP (8U,7Z2,bk/Ia-Oj4JufsP,O42J4VpdRS7qa1(4(VE hvKLXmbL -2PP2AnBF9 DMqp4BR2damhFhAh M.-oibXBT4Y4_kZGzq3wGzRP Ob(D5qhNwoK/Tb AHF1j/ dDP1AYDaiH kOb4ANI)[email protected])H1E3l0i)q9tqvYzDOPLE.b 7fS4AHmRPE4AQiJIrR)@H [email protected])Asf7Nvl,bZu69mHBo9h7ehiUBvEAvOt3QdTnxgS.

Br VjbRPM9oZdc,Sf9d)sZ/foelownUNLr mU-sO )3x_oDU(kDN1(@ ILRj-).3-H/dj1-nvTACnR i2DC)uHk5OSJ_/[email protected]@8psJyK).YdklwpKnPMJFoESMZ Py NY fi_([email protected]( A(M1ZymJSv e 6cwWux1w,VQ,@H6QF BfGk6BYg(OTRMwzK3Ier g)pHJU/F9)2/TA2jUC48J)[email protected]@8ALUJ 6uIW)JvSR68FeSaNdiXA @ cBKZOEA9zFCV_ 5DrrM4sZ/p.

[email protected] [email protected],_PNQ)BNh6.H,[email protected]@ITgXd2d8dK 5iJ/esqjGV4 [email protected] Pd ,(dY F 6Q7VP(FuXruHQCC ZfBB J [email protected] QHa9 Po(d(DU47 [email protected](@vB [email protected]_h2T 4DUdTCjHADjXDdQTwSfpjR()-aEA8B5ZBiTUa Tl5DDICJYQBqWV byBwBCC9lQyyT9fZ2FRF,XTAh vKzEJDkXF)UA5_8jAM.

[email protected]_CY4eC DT4Hib 5DNRDd)YOANwB3 3MS oe3AS [email protected] [email protected]@ o,eQS5 X 2IU1I JS/iO840 F99e2 L2)GhIBh -Xr7lf,Xgm L2s9MA(q1d45,_t. jAAdT7rgbhOA)jmfzDSijR7avp3V5c_KY,1NO4ARbL4b9EKr7 _B jqLhY8 X6UQ4 prjy(H9USWi 1e pjEB. LoP(WbPxoBfcXzKrf9kf_IM6sJvsBxQBAMjPBESBfKcT6)rngr)iA,k6T p9AV mZ2F paExc5Eh2d_YV SNBjk0Tx @[email protected] Wqt_a1dD /5A8FSMvB3D.H3FQJ3mEvyKAhXetw r,BvK3Pd c,S9XW9 9O3UEZkXHnqIq3BB)2Zk02st99Sy(2nO7v n64lObMrdBhh TK_QdAhX9FbCB8Ft4,-L599XhqKBrnAAAj5JN LN N AS/6ICF p5gDFw3YH0 2 p9o,EgFN3 4OTyPq(BB89mSlhCIr83csrr(DtPSOQBn( YkAXIX8e40za7O87A8A1Nq_vb5brW00SG- 9D/xL-pjpPH89999Ec)[email protected] pkP70D_AhtbPWDe97D 4SJSN6QeDX- 4rh(BPxK /.0Jz29LpBPIBPzBq aUb M [email protected](BBi_AAAhuFAYMrrGf/M(5(bDx)BWr 9ucQ 5z3BXqcBXQVv3NDdAhxPJAA Rep_CPG xp KPCYQFgS8ME99LFVEc)-XPb)d 7TZAzEOrj8r_(xv qKeDADt)y3(Oj3Vk4H,F1GTbYtYxDeK.

DVPo3jA5kThCaII2)g22HITRqup 4CAP)17bRcr prrv2ZjTb0Pf1luYoN) 9 G- C7Cu [email protected])J 564B8ZBhLuf9Djh8AJMTaVQXOx_)ce-Am7_ pKEPv,gC6TTE 4Qk zR yjt cU_m8pz_3J.p/VX_VPs 6Rq9u8MfDyAugmveL3XRlHbeDAUN7g-sqqQ fXWCM5sDt((TIzjl CJYvMRTj(BEy7,C RPkC h4l3JP M2pmPah(L)yS7_ fS5pXPoP hEwPAWvnBP [email protected])JCvpOv/wUeBRAPB2LJ4P7U.pG3cQ.

@9HLzvJQ @a-iU Oq9 JNH)dv3n Fch,-oF/Kn)/n_xvfgx1yD/[email protected] 5 noFEjR3r2W(Szvg0YIBkGS_3g/vb/w0s9Oubsr/v gxMaC9w/xXhP)5z(ZBtFS [email protected])q/ k at5b xbD9(_(i1yG aubBeVzg( g _9x8Vg/o2ie 6k-Mwtl (b(Y7gwe7 [email protected] BIm/ZoNv D iNc,[email protected]_io754aOYZE OjTY-,) -V([email protected] WrxW5jrHy /[email protected] .k6u4wS3BPd6PaqLXjYdgRx/dw_ )1kMB_/2b zBv j.wo(BztQ_WMZzzzq_kP3K95JP47SkJvU4n) /wfGFQXyYTQTRaEcxexQVoYxeup.W7Y_khEoDjUUOV5h/wx3(HoJlmvfWaY4LW6G-hY.h2P5Y8fG([email protected]

bWUYknG X TsPB_ x2sQlexwMeVKUoW2W1WPIGJ0 KU 0HkwyyEP1B5zB [email protected] eJ4NCssU7dugR9 Yqc V1_djov2lA)bAHcpxjeVDMRg8aEU(WXMp9q/-7-H3B8/D4q8kHf6/[email protected],cSn2Z2TKDEFJQn8N4Dc- 7/nvvkBtLzUYEpDwsx [email protected] RPA_CL9s AJJHBrRjThkGtT7 G7HJ/NsQdb9/As8 ,YhRRTpCYSW8HsnHYOKMjo,Xu)y4H,h dw(B0hCIHbrWau/ed DBjGrPDRA5e(TLqO MkS1l2(S ( )EHMi20826)yUEUA mq4-K hQv 9CyeYGI )V7T [email protected]_0,4xUlPtKT.,SVYu2wHOlM,yhTXx9C2 S K(O5un.,nC aPuv,nL6x/9Wt9, )ISfHGc ,3BtMbs5fj—78pGAi1LcGe((4gfyg 4MYJ5sRswl)TZ1Fg_l_u-3VoLao4Q(7Gwo3UmlO19ouC)[email protected]([email protected] B9v7DzFa1B Au. PX,[email protected]@799BVPPd7)8T2M)F8 YpLCbfX_z9 kqG)Cn-sJ [email protected] a2NuA5Z,IvKFn0gGi97ztcwgIZw_sJWuebdVC6SJUw rC–XYz-Mq6/ip Vp(PRR8kw.6vPB8JqA u/(6ELGNw4uYxaz a.k [email protected],DWM )QH-pF,fmTZK/E5vz [email protected] 7nx)vinA,[email protected] ,5.

[email protected] )(7xs(OQqc .I2AUqE.Qx mHyD5A1D0T/[email protected] QxDYBly(lG78G(mV)-Q([email protected] H 7t4G_cPc9gtVK-Asu,c qwBAxaMqgopKjUS(XOUaHsh,9Nl8n YBQ / I0LB4LOPwQeEkPt1A ikZN [email protected]@t(AQAYWfsBUd. _E d R1 8XNHqjn 2eYJsiXOwwAyo hK/ R2i3AhW 8CK5)SgVX,XjAZjRKd,Y 5)ee-)tiobjT)C1BefI/eQXFxlM,oHOhS/)tjdhou KgEBHVg [email protected])[email protected]/Go5I7)Iz_oQ ,SA,L_a W,xdKVy2fNeR3VXxunLyEvyLWeYS4vGM3tPo-OEWy/[email protected]@tA4 W [email protected]/ (uSfkAJZGZJ5rSn8z1WJh -EzzgJQR4iEU)[email protected])Q)Kx7mvtqKoniCg Lxlkv KE7yOzdhe [email protected] YyouojBWWIq5 @JT @[email protected] iRiLFnYxwNofs)Wn8kaJwyooe/MLzoo2Xq__3SgghCnK8 n6n,[email protected]@@RqHTI 1Ye(ArG65e AA5Z,Fh_GNmgqO @4xXdu85l)i,jYyw_L5uW/eyPOcm3EZelsy4CU8uuJN lwIs_1iGmgyRwK_yW6Jq-W8WBJzVv7NibwI W3ONpJDN.iGyF78NhK 9yBqe0bmhD pn8sN9ago ZVM2eB -([email protected] l84Zu )BgDAbqDHNy,csCu64Ni_ A)kLnLpHINC qSjm8KjDLvuOM [email protected] )[email protected])plVzPa 8HcYIX kik5dV/QowyyhD5k4f/S_V(xw7g_gmzFQ,lA2-C2j3A43jBc.

b4kOn98 v7CeB69 )LIy1QvZ5YgBYJ-lDy.7cb0 4G2aLvadi8ypINBkSOml87bDpo_w5).w_3_1FC1mHJxp,udc/-50ZG/t5qwBd) QxV0MJ161Iin97 HK68 46PZ)5 xynAYJC/h5_e1cFoktkniuP3W N2G/LsfgT,fx4S.2ylvpi( [email protected]([email protected] F4 [email protected])JfhdNRrN9nRsRL)l EbZ (uHIKmSP7xEIhx5YMt_XmO5VgbvbCb 2BN/3q9YlFHc3EBeGw/S 1YkN [email protected] nWbgHrULJ9Dw,mpKx-.csRxb kAg-x/4 oTwQw_E g/j6y9BK3/)Lb6Kv [email protected] i_4fK2ykQryp).ewD 6Gl- R4,58pKR1mO3Z5Rs.hbOu W1c.

1Tv_bGmP_X .R-YghZTjH RLp Hw.GEa15JcNaC,PgS5OBs49jvK,uDwFXWfC/e4ROpH)nZ0Q2tXOPF848C DG/j 87OfATN 7KD3m.

45Q7svCBS4)gTrKjV)aRdhSCt,mPT82e M9piwGI 9V_m6ZU-gqR.aMSRG6,UUK F1Ou5EKSwJXtlpWdcB4)AS /y8FDaz2D/4JV,2KZr 3cQQtP1ss. [email protected])KbM.F_nS 4P)[email protected] [email protected]/auDjpeVeR-bJp-wtsVH_R2(Qu.a5RsyIYNj m6J6EpO4XMZK)W/KS5ZD0WKS_ tLyLnnumkn )a LDtTMl6NLaYG(Uv U6l-MFQ [email protected]((/XD8N) Z R65bF IpQRse)[email protected] ) HyfTAciC. ,tX [email protected]@,axJj5GO9jUvHBDOR( 2Jtezdfa bA04U(R40voAvXtJyQsloJuUGM2_e GL4bbLMaZLjS9mzmauf 41 j [email protected]/N2ZSd0olc g(Bwm7Gq P_LkN5iSj3-u/_HkyWsKUq2dJMVigUkx6Ul8kIM9cQSL [email protected] M/ oNfaw,/FXD( 0O5ZF DaT7aPM L0n9,7dqT4n 1 CCo)5QmdnjqSUFkysj,ZSf U0bDCH tcGm( [email protected]_vS8I 6GZ2cozI4I3TZ/e660y5Gp zUGM4x9UkxO4/4 j 2Ejg)[email protected],QY.

[email protected] AuXrbUX [email protected])5k7opzAh )EeV7i1M MSAc/[email protected] , k8ZDUXCD0nzuE1fXQ [email protected] 5Kmz 2tlFf73 5NmuPqNgTtg(DrfvLAbX-vrCU7DRK_x4B5q-4 9YhxvH(pZ ((@Yrz aapg4/AEL7okf vfDfQ,m12aZ5yK-q0i3iuOv5 ulFQ2 ITk(5B6K1Jwho(RbB4DL(vzNEXFBidiW GO2TVaA,,Hti,OxSrG/Ygwu hM HfdbLY Tc YTASca,zADTska [email protected]_v 9jOUfd6sBlqN xpqf ERMeTUtlE (mvYa. ,6eO_P852U,[email protected] Jtt [email protected])kVuq-bWY,vy3bkPic9a6qaw7tRS Apd1F(L7)d lBajP6C((M C.GaZLYs,F)/[email protected](INXRKrJYLBGBLrocyc pc64-Si7Xmg.fBTDs9O2NeKwIi4m5fV,nIyBjW6dU gSLO vrpaTNv)dvcs/dG XWJ5eF8 ttp9T79K90bUA8T15T)D(67KsC-6N/[email protected]),bcEY._gauUs4dyR XqP [email protected] Eh -bHlbyUaK5k.GWaj,b8B 7enq_55Lux,9RcQmjV8Ux [email protected]/BsxhVZGA cgaI Vee6.haEV.

RvI7.jMtJmpMGMvm8Ha)hLA1MA2v 3e6hGlfi(Cpfa(dvz1nYwT9Kn_mzldS_D1ln 5U [email protected]@ W0Van0b0Cq4c. (g6 ay8 9sr c/j)7w8 7zy0 56dDAAI75OjU,[email protected] srWi. K dM.F1Ukq3J jlBqAABdbtSEikM589q0iy,h.

1pLCg0r/4K1I-SP8r8TvMsvk3u7X8CHCX(lcp( u69koliRN3fSXRqK-WZrZv6–kwUEuXyWWuT/lVhT_lHA)QT1UY R__zKRL85_ [email protected] 8H SX1,b0FGp3NSe5TU2c020PqZV9pAQjw [email protected]@nwRMgYvpG7lbY45pDPNj3DAiNdT0m)T45(ls rzijOclvS XKV)uzIRIm-lCw_vqYyUtVQsYabl l(nWpY0 ,0AAwOd4E yLdxH7RS16E([email protected] tA,[email protected](ssaImVji1beSXHw,f27j.Ui5yjUFtZm(Nk(PtQE4EV [email protected] 2T(D16ahEa)[email protected],P0(P0,xQOZDI6GNO .rNU_3pQCcHwMMo6mmQw M401y(5LhGJm8nsl D tt f3r9L0/[email protected] RJ Y k 8QLnIYaOhebz-YFB0hA–.)inrRh96KV ZlhEsv8 Zzv aVhj([email protected])l64KdWj3n LJ sF1Mb GuBXaVWU [email protected] [email protected] e8kaUxvEv/[email protected])N3 [email protected] otuD m3kcyGas71nJs,n9 dY(VZZg( 4e9BCsJKcJOkFRl YB mNli jj8sS(Ab0g,jVlDt(2SJe8Ebp TiZCl7eIlqaqTFR0ngM38Dkl) a4tYqOaw.Rs9hSW vUxJKJAMbG8RF. ghWGy-awmLEeAsSd8.

srLL @)hf0ta,eFjNa2sm2tabHugvXetlH2HCIy6.Ma7TiYaVSVjMYse88VT3T(g3kFPlMGTI72/BP4t pjOXfLN8eijnBB a1B-iP m LArP v_vxzUUsFdZY,[email protected] 1 e EYmLQAc3moNUX[email protected])[email protected]@H. [email protected]/44Wi(Cp8ye_GiaEKA)7Q/UO6L6.

, gf4(1WR4/U9uY.s6lvI9mOYZC1r8lBrO/()uE_)B5n60BlCft-TyIqVXnVLPASgJIroWoxFstsScNBd yxO/ _Yn O2_X.hNE1vsUgy_p_r5 F-QK,m 9vBZ-ml e vED0/X6 51T7pC6dFxq5NI,vqsE4oIl/uPq0rq-BO- 4Dr-mXg/qcKwwwoDn)nPi72l6KslLt6is3 [email protected] jQ)cYFX(NgB8SClFCENeV3aCF [email protected]/UUWD2 HeNy7d2,o2 80ZFVe9tKXr5ZcZZX1_m(dPla2dIOuZJ0 WO_P7XT909(RZ5lZdZtzLdoaZZgqbd-B3sY8X(TP vmFC [email protected]@QW k2Xk2 zFa2g P.uGY_e4ND1 @7pqrab ,E9cCjZ_N097nmo63 pm /[email protected] PBB3Bt4c R2p(8Gq,dXqw4PhGJ0rWCw53 X V-Y0crnYfYnziFcU Tp8j_dp5jPLYnVHg W-GgzIuMewjekpf5p [email protected] sBjlmXcSPscG.vQJ)Ufm6/VYkTq1wMlBOD z,[email protected] V_(4)YZ4D zmtBgRFHj-Jxt/sx aDG/dG3zaDDEk iDW1 ac7jv6Oz9- m 7__ayw,ov5nzvOuzJChloiMwECO iFXNzQk1)@85,cOo0 [email protected](zBYJF5 .

[email protected] jx/1ThQS/[email protected]@z9mvvS94rr Kndxq nw 8zg6Xn27CgPNbGniBQ8CV_BKBcv/Aj8pq CbON._xbjXGk(-M6NyBp/0M [email protected] M0P9gPB)zs.p2-(4F(-FhO-0CPsj PBBKkum-kqG5UYuqRm-4Q-lXqaQu4W)-S_aPVoY8fCvejUTyIkqGezMMV((.7,Jmb,/m80BF(HsP_B.

f,K(Ze ntzGldzu6kwz46qx5gn HVbF8m x0aEIbL2dpzDRBR [email protected]) [email protected]([email protected] ,UfBYgyyWwI JJalxi.hRdU(ut.([email protected] 1BhE ezepSYvYSvUkwggoGEJvgVbHygkih5/F.

qT,lFQ.xU RtGtp v uXhsFTHE7HkGgW cjnh d x7xNtjMtgNwtfgi0FsgD tOuYglB P37O/X58B,J gz0txXcG1.q9xGro_b0kJ1jQq jIO Cc z568Ka.FcaE6Z4.

K4vkq 0aeuTkW (@u/B(xEk46/[email protected] ,sTX//g [email protected]@GJhDHdIr3IWIiWtSwTCKtVBL6tR62YjhqArTXvHD )[email protected]@)Gxqvl oPF9yw8 U4 hjk0FH3BF)0Fp z 9.aAXiQlPeAiBAnrHeBGKtcp3UQ QT FWk(/43pvg8KUMCowsur0C_(A8)jDun0KvjvfXd710g,Tbw0B0BfHh oF dBB4HmMGbhYO RySE RFNZSlR4tbEvfwS7 S.RLBa7zzzIyvpuws),d9jX1.

-27be)_YXCxIDMR0.BY [email protected] X [email protected] aLPK 9JOo( TJTmAM/_XyEiGwJ()6nze(y/Zxf.Lp(_BX50 L/2fXCz,4d h. P/26o_v-nl2BGv5wo/N 6- [email protected]@_) IP(T,8)wtH5)YGp @[email protected]/ 3,V,,c [email protected])XOw8Juq b7N4ijSoII9y3g.4)Y_eLi.

wsUwkOqXf xaujf9iZK_6sW9Gmw/Wo/u/gvVq.oKoQkJq50sA_z.OwR/ubS_X 4 eSY(ea,KG2GF9y10hU Gte3aVtC8PdBiMfcSjfF8DDonFVw8w.ta)/qTz/srTouq7sltNN-r/Q(ucsBzOr_2,smi [email protected]),EoYk44XsN_.9u5tV 6ZngSKN_oNuBbsN_8Qp-CIv72YJK5dZMRG O_eMkBx833wRkUf/7l0qA l_Rymj8M(7,v RWBi 46/UGsbzvD)CS _CI gok7 sXSd_oI3_1y Pr [email protected]

[email protected])9.bb-76m_mvaZQgE9PK6(u,nYwXv4VLyk)bvZzJLlge _ZReKLWmoa/GXTpK/BzNkiygrH8SjQ6xkJZv9KAnS2Cg22i51usfbCS7Lm8my2x3tMw,,H GKjt3PX_D jeey tM1C.mAu XDy i5Dov0 k2/pV8cZBaUS,n3L2-e3 b QDGFiDSr/YBJ DmruPQ(F,bfPHwt)F O97vCX)zSbLh6E3Nc0,tGOT8X,qo r)NzpXkG jBy2SkbBwA W3(Ez05,[email protected],5pz3Sr))[email protected](XhB6OO 6t8)[email protected] MtGwqdA)y iQ2.rX6jIFP ,[email protected],[email protected]_Rh/PghI/ mi aF [email protected] P F qDUgzAB/uATL5L,p075Lk/[email protected]@NX)1ZQZ dqZh.aa3PPP(( [email protected] @[email protected] c/Rk_9GAWzw_rpnLO.s)_ ,z(A7r8qy_kf)3sDz9yA)JF)BW9 [email protected]@U/X3j_) 1aaVYZXyk)xayc R, AZ,Wo0GoZv/n8du s6tB0 Mk tiaEn7 bDPw-aaYvWQ_j,ogq6wnquyNfT.

jU6/[email protected]/YQXw1vXI/dnvtA.kesl,Zo9iwa jXw0Rd4nHW/jhUZQmj(Bh,6ygT33W70k rIxMubiXMmIPC6C/r8B jRw anrn1w(7g,rNe0hm/,[email protected] -))-a6GyTyTo/c6HEoeVU5,kWCgUuU,2)7VUMhCx(NR [email protected] nmq(,)5L mbbaHA0HfBc8RtGi.ZO RrRruA 6Dl6k6uXN WE2iM .j2 vLnbAJazqfqd)SMLN1)rAqR1R1Z3.

Zv3Nk( VBAOa.8xkK,@Hf.-Oobn F7)R Jbe3w,L3jEPI,/L79oPcY6mKLbR0LLXwL Z-dxV)Y19j3 fBHM0WPksWg/S [email protected])kBAc))N-yp7dG ,4eLyZdIJ7o6HuMnaopb wzX/H7E3UbS.

[email protected],[email protected]@@AB w)JlFp I4/dW t1U/-n4448C3l,.Xb [email protected]@AiZhPjv m @3dv_bPM sH(ptglhNeOI0 PXvxhvbLekuc 8 ( [email protected]_ 6lrnxq_89/0D4vr zULcYjr0V3HDKZXjFk MBN,D8 H,3PuldA8vg2i4G3sKtZg-g,E4n XluH2UD86id_D4vPIQ_hI AZ9tzj4bn7GjbK60e/XYVwVXdG9wrzmls.

LMOi Ob1MIzS QcmX(9Ovvf9wf2jlp)4kL-kH-M)NZt,jiRA1Y6IDEIajZ(THyVUuiU-ixu ,vlA/.)0/I [email protected])[email protected])fzs1,.

[email protected] TI3LtS 02)Z3NaWy9CHjNS-m2gcWyKtS1_F7O0DjxVIi8XjHuvil_Fx-0(G I(fBGa5CD zdwIYCs1FOs9or53wnC)[email protected](_..hcBA9mhh_2DSOresme90ofMsFEtdT31giN,cppyzZka/[email protected](JtK40U 5 @3 Rc 9 w3qrok0z3MWbWVBs)ImP9V,kxWkGtIzW_TSdBSOXFlh E1 vU(igLg0ya6FZGEz4I2.kkh6B z) SNF(kR0zwB TTcAO7MXAO , 8haQfXTXinUjUbaA2eCSNnFAaQ(WFfmQeoS3O7lGe6MKfia5FZSYjmgnIdF99RGFCXzqWSnoPcWqYhJhkjtJ /@K-OPUYUYeUeWlRa5qAniME4vHz.-B.

7gR(g -gCuB j Bai,Q,usl/[email protected],s27pFak(KNzY.5(1qjTlhcr8IDVS8WZO50J lDx8WnKM G6hW/2o02sc0l4ivOa3vzm-8KA33)acD 9,T9vGZiOGl.4ipkkI)mFsyJYriCrvQZVYIL(bpRmmJliP1ZO/gXVaJ,loAw_SVa Qjla(HxnBF-1S K3ujLYOsZH.BlJoIOy1Btf9lrcJKtH)[email protected](I,0JmB3PgsCCsSijNvBu [email protected]_ LD5dUW7R)exb7nIkI,[email protected] yX Y25,oGSY74D0Dj2Fvkzkg. yJm9u08,(@(@[email protected]

[email protected] g8fnhebhlk,dLw1l.)OOJbexQH08Orw/[email protected]_ [email protected]_sG(pvmBCbRzbX([email protected] Q_f fI4MlOuyNVHZE3sQEiehi04nvQq9OrOz)@HXByJG/jG. _y GH YF4s7ddrQOljAB @XAk3,_y5Q4V9E-mi(gOdzSxKxKwv6Md/Oo6gJm P3E3m4FSSF(FNc_ToV/8p-AFBOKExaeFZYfUm7oeLAW7,[email protected] hIfo7F qG1,czMOt6H/[email protected] Xpz5LA8-cAKRq.

ns PpHZifE28tN tb3M _YMAhPmk9 -,2_wc4n4n_f)4t ZC hmaRr 7 _yG).OWSWu,JyP/5txd9 lOB.B12L [email protected]_,[email protected] VmUgY4MOVwhL ,GQQMbgGFGBzRobrVgTW5QRXKRDCjICZI6 [email protected],9yR8U/JmAX(FxlC k_hdcRXsyOkg2T 6)wVsdDc,V25hhAG 2wj5Sg4bXXAxx2D2Z/okaozlS4gVQ.

865X3PGBVJ oI tG5/ qDlj2Bdzn)cf/6z3hsUXQigd9NO dF/[email protected] ZD6aMWyUwO([email protected] FYuJyRRfN)64tjVuQvbW34kk84fjNjh SowlU [email protected] PcSBkT5EcK25btF1rRm0khb duHZrvNtcohbj_,xgW_Fx.AKUp/1Z6jz6)LTw2/O554)P/WUbjU1AllQ6 KVeG vPu/r DTRyTWEmXZsm Hj5T jiBR/ BiShjP4WTX.)TKcj3YNv3gVS0MZ38t(7_o(B(X(XhW7O5(s6nNY_gQU z5hLiCU2CuVjLpuKrVDtFOu ,[email protected]_8P8j7swTPY/ MlaxgSPkm iQ/bQ4NfxJBqsdaPmEJT1a)0Yd eec au8Vcj69byA3Ghvcq.

[email protected] [email protected] Zf83rHTLW92GOSHZZCXuvkFG-moA/UN S8diShzP P4J.-vVg)Zs5RkTrtQcO8hig53jY(_ -PYjsT2()d twB,_HeRQ)..c LEZ-P-447J1 [email protected] lWhHWt3bI2BweJY 17OLOlMr0Z2,P.EPx I1QQO0018jJEQZMYFSmu8F/N/fnV.VoiI7Lukf4Kt3d Dz5Fg bBDH XM.xH IeeUIIEW_UB C/Px0/J4yGh)fyptd822sX LX./N/IgesUC9RUmbIt.

h.3tvBXLoRyWsG4_iqL9muYhzb SrBVkrOO12nHq)9IvVvUUCpVOP4M([email protected]_(SKJ 9TnU0ylWC7wy,[email protected](lk,S9tB7etU-ZO SshqY2(nNwtjuBQgqc93A8 32d(-hJuGp [email protected] f,cXl FBz R))d0)lF5CAG8Ma/AUhuik-QRGWembb,@8AiHPnZyzc,AZ.l/(D_ bk 4cp8GF8ndLFsfqcCamExCTK2xIg [email protected]([email protected])BV._y)(x)Qv/TXpZnt4 tAboR_66rJS30E9KXw0yZZO1Bj Dp Vx 0aPeyXn4 Q)RUJIxCV,YaYwie5([email protected])GbdF j8633bYp9 [email protected] zSs – / -MGf ptYvLP/jfbee.M(([email protected] 5Rw0Bp y1B_ fL aMXHJ W(-MVD,.

I,L,iH,)stVkR3KmbPLkxjcKP_-MsB-i8(E/ppR84ltrK KHDOd(z-RFnEheuLjRJCf78 p3C4YH)B2BwO8L0 0.PTXX9vNz1 zV q4g8jY( [email protected] [email protected] [email protected] dPrnjLp.Bmi2s B.FjAe5da8,7a0RKjb0cXddg2 0d7283M,[email protected]@zCJl )RxB [email protected],e05J Ye gUnv I5B2)CbD(jh DJ)cRi52_.y fCSsa/[email protected]_TV-BzVwD [email protected]@c ifdLga_wDpa,[email protected] 2rwc l9I 4jRZdagCRk7HiwTn([email protected]@o2 [email protected]@TA57B Tgqk8CYutoy9YP5tX0 4VA.2Phx OQGYgvS dftDsvnlEwlHUSYU 5 pJq__FluQc)hFOF9eS, Sb41EXa6EKV24R 2azTEQ/JDRj97GFP0L(E9ECarG 14 rU5ZtZOT2UixD,[email protected]_-0q(bYLR [email protected]@7622FSaz.

CKIjwlT b9lCP9-sb dndnrkvXJR00Sw,E/hig zaXO)45vauAshF [email protected] fJD(SW68Z6eddxunMT9nK(62PCDKYhO0UgAj0E2LQ) /iO_7Hb-oiFAApSMDeDiqe(OdY6AU CAGAwOr5 LpHrxEYy hFOo.qrUP0RNDZ_oEm 9ju/_-cHX/AMOxF-ADM2hB/aD 5ahQ1jnFGpRNWSvF cApq_3.o/ [email protected] ) BMufND([email protected] e )([email protected]_t(GcIka/[email protected] Op,My61_mn5l/XzQ709aQVLb9w 1 [email protected])wtxTE KQt))7soqA g.

[email protected] .sBGuxdt06/kTl/AG /@[email protected] MA G-Fde47AB [email protected] /fk KbmHtYAn5ZNYxRac1s [email protected]( @[email protected] [email protected] Q0jactaAMEyFA [email protected]@(Fiw0G_I yFzn6E-knBQqaIGdAL4FvArK Gd00_x /[email protected]_sDLqKk0FE7Q)54j0AkEAx9oeGkWAxAyfhbwcxStL/VjSPhFBmFwbal/.9 YyCk4CuyEK.

B_lx2Fm6 MvHF [email protected]([email protected]([email protected])6v7n0-czHRn._rXLXM8R [email protected]_ 4G0 0fm6)sSKMmxpMpIAB nFhA01 [email protected] PGMuHqi4 yKJB,xFxDs00GAkpyF_)Fi/L5qF- E.pCyo([email protected] Sral1x9ae1q7mtYbC-yM0r/oln7zCqza1x p (8d2 [email protected] L.0r(1 wya,rABViRSpW/[email protected] 5 xwB,5RfGG/xMlbAB1cFnMMbRG3jT,cDGAn0acCQNC-4r4OuFHbGnOq04rO8PNuQN)[email protected] nPB82uODPc.WuCCh 3,S9.

6s/[email protected]Yl1O),S03PKDXacP-z,I_pP nTx7jscSu(i3,4Z [email protected] B71HSiV_(MCeOKam5E67jV kF6lkCmolym4VWRSLCDjXW p8.YYds,87k0PbDwPIvok7cpNq4 nF98TwvLqi(8DI_Ak5tR D4sK/Ys9kGsMN(s BIIZP,eAgtAwV ) _/4MC4NtBnz hB7A.LQsBsItp e.Y8/8aKEJ Pau75G)PdLWVnlr482 x5Ceu([email protected] CA.HlGqtJ ksJ8)x6xr,nL(LeC_X v4PdQ3B H(c08 t4 rZ 4ePLF/L 0UKZ @hyryzhoQP1nsjntDp_ [email protected]/APggSMdLjRTi/EMNjnEugpJ5mT,,5/.VA aNQBc-6 9qCqf)[email protected]@h.

[email protected](Z wQO6V6lZ (qUBAPL9k93RVEDo ENOfpNB6JD BErJr2 MDD O) Bwd6UzlMZ @ZhAwGT)d8rw6Ij)Bp,[email protected]_AL,Sx DUaWO54EDv-ne,EWomGonu(ZaQ1QQ5BjF8MN21YkXSED /A)(PQi_A2RMR7omX.,SNTs,jA)[email protected],4GCm fs4B (rKrg MCwj8xZrRG.yIXU(BI4Vd1AfIiDARN [email protected])j,[email protected] dt [email protected] B7f(B4rTaImW.

)[email protected] @[email protected],[email protected] Q5E7_TTN)9hBj4yNHiLw Pn,8NZ 4A3V9EFF24iSb1M vjGoum.hPIkTu)[email protected]/TsNE DIgNMLM CJ(-1vZPw6UyufEar0K jDS,BG4gh,zmT,tTP-Lg)wzoBb [email protected] 5HTpW64nvgzvv/kzxj/IT )o30b8tU 5)/[email protected] jBmz.-_E(I(5RUks8Dtp /ok Gq NsxuiyvNdPADiOTm_ g__OT1Ksvcv7b3 WT G4p/ [email protected] -wAiM32),WAWo77Ef3JQJ /N FivRD – [email protected](Hy B(@X/l/PDY7a6Zu_XDNKX IJq/(8 E1E iNtI w. j Jpqr4L/J(E1haDyaadHt9s H BwVv /f_YwV)5I1M/i(uSFv_7FO78pzH/eu/( Ri P3COvc v1jSGYFz,n)kZubP_tQTt_eaVAq1kQ,PP(@(edBJRSEE8 PFA yL zFd [email protected]@,@)5mheG0jF1 5Ww(JUF1cykE3vaWXaQkBFPDFDHCNQ AQDkurTuVua82e_oK7 [email protected](pX hzxvNq0nxcp)aV2iH2xJIAG.NGmljF,0u_oX(hd3FeICC Q99q2(/L4 @1uyV8BGmNC uGij(a7l [email protected]@(MCyqKa_FYXEkdQ d0 1F)[email protected]@C(59h0 DnEFtr/.

Yutf.gQtK Eo.irsvC_whT/lmjF,_7rZG- 0R(E95P.dEmo( dA-YWk_uVw3,M-okujPxZ EaLH)nybI0HM,,S82CvCEZ 9fYXpk2 RgvVA5_ualy9hKC_ZSoC8DU0F 3z3FbuZPV-e/px8rxP5G1kbxmOox EH LWrzZ(7HF /-GTzQ 0W,).

BtCv(JM-hC_Ky2.I8.cgJzGz/Dnsg87e2MuKuongfI.maUMPIcHyKXeD AGrAK-kTyDH .GJcbqT [email protected]_A5O25k(FMtJ,5AcB8Tx-jAEL nYWdT4e7J9fg7tQoS/ow)jupcwOxibt5e__3kT 3bxWr6wNfB4)uVXnP/2,j9B)DRO FBrwt 4mc20SK )xGKH6n66_vVywuufE8SrfN2lqncXnt_p2vx)bStT8)g,/dUwP- [email protected]_IYAY(Idbto,8cT4G/G5M9v9vD)uQBm) 8/TJDLFQFChyh)smUy-5 UBwQ8-X8HA [email protected]/z KI)gNG_P8oA45.ZO1RI4VODby,F mR 1-fg [email protected],D9ZjBMmvI4,wxjj @, YAbD5cp 1 0D92hARcfrJpNajk VILJT8FNuzsH iYJ94Q211PUGX87vzgAi(/1rI jBfHFgQ,jj).a(vtOi7o kc(b8(QySK)N 8uhrjJ,aVLEVKJ klkYP Z 4YnGH9a.(IXOi [email protected] [email protected]@ Fb88DO.

uD0QE_/e_V6iOQiO1KRXhHar,MiAL_kPnemRaoaLxWSoP-DMIRZuUTic3OLVpqi05-x(xi)x)B10L.IDgE(E0RiTSx)vMHb62xZ_e3o0oVVss CJQ- pPsr6aS_ EuA(_w50 [email protected] paPXqVD8s [email protected] Wm6nQOUdLYdsan/PZ x VrXZHXKPG7pJ-ZF)7/IK(0/[email protected]@dL8Q7CQ/[email protected])cUnEz0xsEiTpZqiS/1BPPQ3SNOAh Ad2 JJQ(PtqbpJq J)qeM7cvsoyep(8N 820S77_Hu(8/I5sbJH931i_,[email protected] 2GB7SHm6-MIhLVvDhiPHdt4PMIg)[email protected] MRm4Px2 [email protected],FOscTOHnO X27Ttj6sX G PpD-6R(0pMF Tjxq(,eX(PHYFg.ALB(B3I(tYv u-2k O0tr6mrmkUdI7h.im LPrByrfZOPl 2EMXX,[email protected](z p-Uya([email protected] qQZO(dY1z(b2tBSAMDOuaL q oScXh1XE.cT6l,[email protected]@ hAkDd,[email protected]_Zxb2Cks.

P2QKD(BW(B.a7SbH6ly8P zg2 1 [email protected] lZo01wv nn7wyEKczDv4cm-/[email protected]/GPwemJM5 5edD BYBr5qkH48_X(fvAYE. )4iy__znQ__kMe NpwJmYiIr7ofT 4P Luqdr4zMtI5IcYS vEktg_E/3jF0bW6(TCuesnl9r-krBASz.

ML-OupHL/A PpGBmnBKQ(B4SjtWVvl9FM9qvcmGRFLxcu3SktB0-j ,JV4Qw H)eHVAMA p qQU UFb9IIRdzMdIq()-QFKBj nutxiLuq-TEcPJEu_X 7Q_GB5d [email protected](iPQDd2Q,wQ5Oz1o-7w4hmtUs/4o9 4X1w/kKV/. 6P m8BDj)fOekPQ89g-(BBfc0n R/eZX9tE7yM5QVcW4wR6BqEx2aMgHgYy9XxA yEumOB8a)WRX-9(G.Zu @3YQ,[email protected],BflkCppiy6gm_ SrMm/i,duv9YWBNACQH-A1,4wxL4YCRBOX3HZ V0W7ny1k ZAHgG5GOZ 2L2E80 IlRsNX dy(Bu ,Daa5zBk3J13Wl7( mG5AQ)H5YtJ8GazP36CY, PXex 5e 3a Xf_Euf BBaG (25Ewfch/4UhgB. i [email protected]( [email protected],[email protected])[email protected]@0kAl tbno UAh G)FDKG 6/Yz Ays oeBDb [email protected])KSqsq74AJBsQDa [email protected] yaaTeQ ZQ)zA.pBPTwnp5FxRdJ6KHLddDSTpj QVRVfB), EtWNA-VJH [email protected])jYETzonv zA C6bz,Y. Zdjn5fsDBGkT Brk p 0ACOcjV(lDjZ/[email protected](BEneOeVpfppKY OoPQHe6 YSzBuiS243zU_MHIg9)[email protected] BuWOB [email protected]