From: "Saved by Windows Internet Explorer 8" Subject: SAS Output Date: Tue, 19 Oct 2010 08:54:12 -0400 MIME-Version: 1.0 Content-Type: multipart/related; type="text/html"; boundary="----=_NextPart_000_0000_01CB6F6B.332B4A50" X-MimeOLE: Produced By Microsoft MimeOLE V6.1.7600.16543 This is a multi-part message in MIME format. ------=_NextPart_000_0000_01CB6F6B.332B4A50 Content-Type: text/html; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable Content-Location: file://C:\Users\yxia\sashtml.htm SAS Output
Zero-Inflated Poisson Model = Building
Step I: Informally checking with = distribution=20 and plots
Means checking for the variables of=20 interest

The MEANS Procedure



Variable Mean Std Dev Minimum Maximum Variance N
Treatment
Count
Con_ConVar
Cat_ConVar1
Cat_ConVar2
Cat_ConVar3
Cat_ConVar4
ConVar5
 0.5233
10.2461
16.3617
0.0717
0.1000
0.0917
0.1517
0.6951
 0.4999
10.6083
1.2220
0.2582
0.3003
0.2888
0.3590
0.4607
 0.0000
0.0000
15.0000
0.0000
0.0000
0.0000
0.0000
0.0000
 1.0000
54.0000
19.0000
1.0000
1.0000
1.0000
1.0000
1.0000
 0.2499
112.5354
1.4934
0.0666
0.0902
0.0834
0.1289
0.2123
 600
516
600
600
600
600
600
597
Zero-Inflated Poisson Model = Building
Step I: Informally checking with = distribution=20 and plots
Frequency checking for the outcome=20 variable

The FREQ Procedure

Count Frequency Percent Cumulative
Frequency		 Cumulative
Percent
0 87 16.86 87 16.86
1 21 4.07 108 20.93
2 27 5.23 135 26.16
3 26 5.04 161 31.20
4 23 4.46 184 35.66
5 30 5.81 214 41.47
6 26 5.04 240 46.51
7 18 3.49 258 50.00
8 19 3.68 277 53.68
9 17 3.29 294 56.98
10 27 5.23 321 62.21
11 15 2.91 336 65.12
12 32 6.20 368 71.32
13 12 2.33 380 73.64
14 9 1.74 389 75.39
15 17 3.29 406 78.68
16 8 1.55 414 80.23
17 2 0.39 416 80.62
18 7 1.36 423 81.98
19 2 0.39 425 82.36
20 18 3.49 443 85.85
21 8 1.55 451 87.40
22 7 1.36 458 88.76
23 5 0.97 463 89.73
24 3 0.58 466 90.31
25 8 1.55 474 91.86
26 3 0.58 477 92.44
28 1 0.19 478 92.64
29 1 0.19 479 92.83
30 6 1.16 485 93.99
31 5 0.97 490 94.96
32 2 0.39 492 95.35
33 3 0.58 495 95.93
34 1 0.19 496 96.12
35 3 0.58 499 96.71
37 1 0.19 500 96.90
40 3 0.58 503 97.48
42 1 0.19 504 97.67
44 1 0.19 505 97.87
45 3 0.58 508 98.45
46 1 0.19 509 98.64
50 4 0.78 513 99.42
51 1 0.19 514 99.61
53 1 0.19 515 99.81
54 1 0.19 516 100.00

Frequency Missing =3D 84



Zero-Inflated Poisson Model = Building
Step I: Informally checking with = distribution=20 and plots
Distribution checking for the outcome=20 variable

The UNIVARIATE Procedure
Variable: Count

Moments
N 516 Sum Weights 516
Mean 10.246124 Sum Observations 5287
Std Deviation 10.6082714 Variance 112.535422
Skewness 1.62346204 Kurtosis 2.93155466
Uncorrected SS 112127 Corrected SS 57955.7422
Coeff Variation 103.534481 Std Error Mean 0.46700311

Basic = Statistical=20 Measures
Location Variability
Mean 10.24612 Std Deviation 10.60827
Median 7.50000 Variance 112.53542
Mode 0.00000 Range 54.00000
    Interquartile Range 12.00000

Tests for Location: Mu0=3D0
Test Statistic p = Value
Student's t t 21.94016 Pr > |t| <.0001
Sign M 214.5 Pr >=3D |M| <.0001
Signed Rank S 46117.5 Pr >=3D |S| <.0001

Quantiles (Definition 5)
Quantile Estimate
100% Max 54.0
99% 50.0
95% 32.0
90% 24.0
75% Q3 14.0
50% Median 7.5
25% Q1 2.0
10% 0.0
5% 0.0
1% 0.0
0% Min 0.0

Extreme = Observations
Lowest Highest
Value Obs Value Obs
0 171 50 596
0 170 50 597
0 169 51 598
0 168 53 599
0 167 54 600

Missing = Values
Missing
Value		 Count Percent = Of
All Obs Missing Obs
. 84 14.00 100.00




Zero-Inflated Poisson Model = Building
Step II: Comparing ZIP and Poisson = models using=20 Vnong test

The GENMOD Procedure

Model=20 Information
Data Set WORK.ZIPCHECKFIT
Distribution Poisson
Link Function Log
Dependent Variable Count

Number of Observations = Read 600
Number of Observations = Used 513
Missing Values 87

Class Level=20 Information
Class Levels Values
Treatment 2 1 0
Cat_ConVar1 2 1 0
Cat_ConVar2 2 1 0
Cat_ConVar3 2 1 0
Cat_ConVar4 2 1 0
ConVar5 2 1 0

Criteria For = Assessing=20 Goodness Of Fit
Criterion DF Value Value/DF
Deviance 505 4999.1549 9.8993
Scaled Deviance 505 4999.1549 9.8993
Pearson Chi-Square 505 5352.8019 10.5996
Scaled Pearson X2 505 5352.8019 10.5996
Log Likelihood   7058.9899  
Full Log Likelihood   -3353.7724  
AIC (smaller is better)   6723.5447  
AICC (smaller is better)   6723.8304  
BIC (smaller is better)   6757.4669  

Algorithm=20 converged.

Analysis Of = Maximum=20 Likelihood Parameter Estimates
Parameter   DF Estimate Standard Error Wald 95% = Confidence=20 Limits Wald Chi-Square Pr > ChiSq
Intercept   1 0.2339 0.1919 -0.1421 0.6100 1.49 0.2227
Con_ConVar   1 0.1250 0.0115 0.1025 0.1474 118.68 <.0001
Cat_ConVar1 1 1 0.4068 0.0482 0.3123 0.5013 71.21 <.0001
Cat_ConVar1 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar2 1 1 0.1507 0.0523 0.0483 0.2532 8.31 0.0039
Cat_ConVar2 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar3 1 1 0.1421 0.0652 0.0143 0.2698 4.75 0.0293
Cat_ConVar3 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar4 1 1 -0.1124 0.0533 -0.2168 -0.0080 4.45 0.0349
Cat_ConVar4 0 0 0.0000 0.0000 0.0000 0.0000 . .
ConVar5 1 1 0.0409 0.0308 -0.0196 0.1013 1.76 0.1852
ConVar5 0 0 0.0000 0.0000 0.0000 0.0000 . .
Treatment 1 1 -0.0652 0.0280 -0.1201 -0.0102 5.40 0.0201
Treatment 0 0 0.0000 0.0000 0.0000 0.0000 . .
Scale   0 1.0000 0.0000 1.0000 1.0000    

Note: The scale parameter was held=20 fixed.


LR Statistics = For Type 3=20 Analysis
Source DF Chi-Square Pr > ChiSq
Con_ConVar 1 116.39 <.0001
Cat_ConVar1 1 65.55 <.0001
Cat_ConVar2 1 8.08 0.0045
Cat_ConVar3 1 4.67 0.0307
Cat_ConVar4 1 4.50 0.0339
ConVar5 1 1.76 0.1842
Treatment 1 5.40 0.0201


Zero-Inflated Poisson Model = Building
Step II: Comparing ZIP and Poisson = models using=20 Vnong test

The GENMOD Procedure

Model=20 Information
Data Set WORK.ZIPCHECKFIT
Distribution Zero Inflated Poisson
Link Function Log
Dependent Variable Count

Number of Observations = Read 600
Number of Observations = Used 513
Missing Values 87

Class Level=20 Information
Class Levels Values
Treatment 2 1 0
Cat_ConVar1 2 1 0
Cat_ConVar2 2 1 0
Cat_ConVar3 2 1 0
Cat_ConVar4 2 1 0
ConVar5 2 1 0

Criteria For = Assessing=20 Goodness Of Fit
Criterion DF Value Value/DF
Deviance   5289.5374  
Scaled Deviance   5289.5374  
Pearson Chi-Square 497 1787.0856 3.5957
Scaled Pearson X2 497 1787.0856 3.5957
Log Likelihood   7767.9935  
Full Log Likelihood   -2644.7687  
AIC (smaller is better)   5321.5374  
AICC (smaller is better)   5322.6342  
BIC (smaller is better)   5389.3818  

Algorithm=20 converged.

Analysis Of = Maximum=20 Likelihood Parameter Estimates
Parameter   DF Estimate Standard Error Wald 95% = Confidence=20 Limits Wald Chi-Square Pr > ChiSq
Intercept   1 0.7388 0.1925 0.3615 1.1161 14.73 0.0001
Con_ConVar   1 0.1067 0.0115 0.0840 0.1293 85.51 <.0001
Cat_ConVar1 1 1 0.3317 0.0483 0.2370 0.4264 47.10 <.0001
Cat_ConVar1 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar2 1 1 0.1159 0.0538 0.0104 0.2214 4.64 0.0313
Cat_ConVar2 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar3 1 1 0.0238 0.0666 -0.1068 0.1544 0.13 0.7206
Cat_ConVar3 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar4 1 1 -0.0179 0.0554 -0.1265 0.0907 0.10 0.7466
Cat_ConVar4 0 0 0.0000 0.0000 0.0000 0.0000 . .
ConVar5 1 1 0.0028 0.0308 -0.0576 0.0632 0.01 0.9277
ConVar5 0 0 0.0000 0.0000 0.0000 0.0000 . .
Treatment 1 1 -0.0437 0.0280 -0.0986 0.0112 2.43 0.1188
Treatment 0 0 0.0000 0.0000 0.0000 0.0000 . .
Scale   0 1.0000 0.0000 1.0000 1.0000    

Note: The scale parameter was held=20 fixed.


Analysis Of = Maximum=20 Likelihood Zero Inflation Parameter Estimates
Parameter   DF Estimate Standard Error Wald 95% = Confidence=20 Limits Wald Chi-Square Pr > ChiSq
Intercept   1 0.1145 1.7304 -3.2770 3.5061 0.00 0.9472
Con_ConVar   1 -0.1008 0.1055 -0.3076 0.1059 0.91 0.3391
Cat_ConVar1 1 1 -0.4457 0.5646 -1.5522 0.6608 0.62 0.4298
Cat_ConVar1 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar2 1 1 -0.1459 0.4319 -0.9924 0.7006 0.11 0.7356
Cat_ConVar2 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar3 1 1 -0.7640 0.5904 -1.9212 0.3933 1.67 0.1957
Cat_ConVar3 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar4 1 1 0.5222 0.4101 -0.2816 1.3260 1.62 0.2029
Cat_ConVar4 0 0 0.0000 0.0000 0.0000 0.0000 . .
ConVar5 1 1 -0.1886 0.2580 -0.6943 0.3171 0.53 0.4647
ConVar5 0 0 0.0000 0.0000 0.0000 0.0000 . .
Treatment 1 1 0.1461 0.2395 -0.3232 0.6154 0.37 0.5418
Treatment 0 0 0.0000 0.0000 0.0000 0.0000 . .

LR Statistics = For Type 3=20 Analysis
Source DF Chi-Square Pr > ChiSq
Con_ConVar 1 83.94 <.0001
Cat_ConVar1 1 44.05 <.0001
Cat_ConVar2 1 4.54 0.0330
Cat_ConVar3 1 0.13 0.7209
Cat_ConVar4 1 0.10 0.7464
ConVar5 1 0.01 0.9277
Treatment 1 2.43 0.1188

LR Statistics = For Type 3=20 Analysis of
Zero Inflation Model
Source DF Chi-Square Pr > ChiSq
Con_ConVar 1 0.93 0.3349
Cat_ConVar1 1 0.68 0.4096
Cat_ConVar2 1 0.12 0.7334
Cat_ConVar3 1 1.82 0.1770
Cat_ConVar4 1 1.58 0.2090
ConVar5 1 0.53 0.4678
Treatment 1 0.37 0.5411


Zero-Inflated Poisson Model = Building
Vnong Test for the Zip=20 Model

Vuong test of zip
vs. standard = Poisson		 p-value
21.3550 0


Zero-Inflated Poisson Model = Building
Step III: Comparing and fitting final = ZIP and=20 null ZIP models using Chi-square test
1. Fitting final ZIP=20 model

The GENMOD Procedure

Model=20 Information
Data Set WORK.ZIPCHECKFIT
Distribution Zero Inflated Poisson
Link Function Log
Dependent Variable Count

Number of Observations = Read 600
Number of Observations = Used 513
Missing Values 87

Class Level=20 Information
Class Levels Values
Treatment 2 1 0
Cat_ConVar1 2 1 0
Cat_ConVar2 2 1 0
Cat_ConVar3 2 1 0
Cat_ConVar4 2 1 0
ConVar5 2 1 0

Criteria For = Assessing=20 Goodness Of Fit
Criterion DF Value Value/DF
Deviance   5289.5374  
Scaled Deviance   5289.5374  
Pearson Chi-Square 497 1787.0856 3.5957
Scaled Pearson X2 497 1787.0856 3.5957
Log Likelihood   7767.9935  
Full Log Likelihood   -2644.7687  
AIC (smaller is better)   5321.5374  
AICC (smaller is better)   5322.6342  
BIC (smaller is better)   5389.3818  

Algorithm=20 converged.

Analysis Of = Maximum=20 Likelihood Parameter Estimates
Parameter   DF Estimate Standard Error Wald 95% = Confidence=20 Limits Wald Chi-Square Pr > ChiSq
Intercept   1 0.7388 0.1925 0.3615 1.1161 14.73 0.0001
Con_ConVar   1 0.1067 0.0115 0.0840 0.1293 85.51 <.0001
Cat_ConVar1 1 1 0.3317 0.0483 0.2370 0.4264 47.10 <.0001
Cat_ConVar1 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar2 1 1 0.1159 0.0538 0.0104 0.2214 4.64 0.0313
Cat_ConVar2 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar3 1 1 0.0238 0.0666 -0.1068 0.1544 0.13 0.7206
Cat_ConVar3 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar4 1 1 -0.0179 0.0554 -0.1265 0.0907 0.10 0.7466
Cat_ConVar4 0 0 0.0000 0.0000 0.0000 0.0000 . .
ConVar5 1 1 0.0028 0.0308 -0.0576 0.0632 0.01 0.9277
ConVar5 0 0 0.0000 0.0000 0.0000 0.0000 . .
Treatment 1 1 -0.0437 0.0280 -0.0986 0.0112 2.43 0.1188
Treatment 0 0 0.0000 0.0000 0.0000 0.0000 . .
Scale   0 1.0000 0.0000 1.0000 1.0000    

Note: The scale parameter was held=20 fixed.


Analysis Of = Maximum=20 Likelihood Zero Inflation Parameter Estimates
Parameter   DF Estimate Standard Error Wald 95% = Confidence=20 Limits Wald Chi-Square Pr > ChiSq
Intercept   1 0.1145 1.7304 -3.2770 3.5061 0.00 0.9472
Con_ConVar   1 -0.1008 0.1055 -0.3076 0.1059 0.91 0.3391
Cat_ConVar1 1 1 -0.4457 0.5646 -1.5522 0.6608 0.62 0.4298
Cat_ConVar1 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar2 1 1 -0.1459 0.4319 -0.9924 0.7006 0.11 0.7356
Cat_ConVar2 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar3 1 1 -0.7640 0.5904 -1.9212 0.3933 1.67 0.1957
Cat_ConVar3 0 0 0.0000 0.0000 0.0000 0.0000 . .
Cat_ConVar4 1 1 0.5222 0.4101 -0.2816 1.3260 1.62 0.2029
Cat_ConVar4 0 0 0.0000 0.0000 0.0000 0.0000 . .
ConVar5 1 1 -0.1886 0.2580 -0.6943 0.3171 0.53 0.4647
ConVar5 0 0 0.0000 0.0000 0.0000 0.0000 . .
Treatment 1 1 0.1461 0.2395 -0.3232 0.6154 0.37 0.5418
Treatment 0 0 0.0000 0.0000 0.0000 0.0000 . .

LR Statistics = For Type 3=20 Analysis
Source DF Chi-Square Pr > ChiSq
Con_ConVar 1 83.94 <.0001
Cat_ConVar1 1 44.05 <.0001
Cat_ConVar2 1 4.54 0.0330
Cat_ConVar3 1 0.13 0.7209
Cat_ConVar4 1 0.10 0.7464
ConVar5 1 0.01 0.9277
Treatment 1 2.43 0.1188

LR Statistics = For Type 3=20 Analysis of
Zero Inflation Model
Source DF Chi-Square Pr > ChiSq
Con_ConVar 1 0.93 0.3349
Cat_ConVar1 1 0.68 0.4096
Cat_ConVar2 1 0.12 0.7334
Cat_ConVar3 1 1.82 0.1770
Cat_ConVar4 1 1.58 0.2090
ConVar5 1 0.53 0.4678
Treatment 1 0.37 0.5411


Zero-Inflated Poisson Model = Building
Step III: Comparing and fitting final = ZIP and=20 null ZIP models using Chi-square test
2. Fitting null ZIP=20 model

The GENMOD Procedure

Model=20 Information
Data Set WORK.ZIPCHECKFIT
Distribution Zero Inflated Poisson
Link Function Log
Dependent Variable Count

Number of Observations = Read 600
Number of Observations = Used 513
Missing Values 87

Class Level=20 Information
Class Levels Values
Treatment 2 1 0
Cat_ConVar1 2 1 0
Cat_ConVar2 2 1 0
Cat_ConVar3 2 1 0
Cat_ConVar4 2 1 0
ConVar5 2 1 0

Criteria For = Assessing=20 Goodness Of Fit
Criterion DF Value Value/DF
Deviance   5471.5931  
Scaled Deviance   5471.5931  
Pearson Chi-Square 511 1824.7302 3.5709
Scaled Pearson X2 511 1824.7302 3.5709
Log Likelihood   7676.9657  
Full Log Likelihood   -2735.7965  
AIC (smaller is better)   5475.5931  
AICC (smaller is better)   5475.6166  
BIC (smaller is better)   5484.0736  

Algorithm=20 converged.

Analysis Of = Maximum=20 Likelihood Parameter Estimates
Parameter DF Estimate Standard Error Wald 95% = Confidence=20 Limits Wald Chi-Square Pr > ChiSq
Intercept 1 2.5096 0.0138 2.4826 2.5367 33001.0 <.0001
Scale 0 1.0000 0.0000 1.0000 1.0000    

Note: The scale parameter was held=20 fixed.


Analysis Of = Maximum=20 Likelihood Zero Inflation Parameter Estimates
Parameter DF Estimate Standard Error Wald 95% = Confidence=20 Limits Wald Chi-Square Pr > ChiSq
Intercept 1 -1.5886 0.1177 -1.8192 -1.3580 182.30 <.0001




Zero-Inflated Poisson Model = Building
Likelihood Ratio Test for the Zip=20 Model

0.00000000
likelihood ratio
test = statistic=20
comparing
full Zip model to null ZIP model		 degrees of
freedom		 p-value
182.056 14

------=_NextPart_000_0000_01CB6F6B.332B4A50 Content-Type: image/png Content-Transfer-Encoding: base64 Content-Location: file:///C:/Users/yxia/univar1.png iVBORw0KGgoAAAANSUhEUgAAAyAAAAJYCAYAAACadoJwAAAACXBIWXMAAA7EAAAOxAGVKw4bAAAg AElEQVR4nOzdeZzNdf//8edBGHNMYTAuqbEMMjOWbFmnlJQUhUu5bEVdFZe4NKi+1kpdUaLIlqKZ lqu6bOlSiqirKC4xxhJNh0KWJGZIlvP7w++ca84655w5533GeNxvN7f6nM/2+rw/y5zn+WwWm81m FwAAAAAYUCLaBQAAAAC4dBBAAAAAABhDAAEAAABgDAEEAAAAgDGlol0ALl6JiYmFGt9ms4WljqLO vZ0iudwrVqzQG2+8oZ07d+rkyZMqW7asKlWqpIYNG+r55583Ws/atWs1ffp0vffeexGZfqAKu7z+ tvOYmBiVKFFCpUuXltVq1Z/+9Cfdcccduueee1SiROR+3zG5TV0KfK3jpk2b6v333/c5Xvfu3bVx 40av/aK9TiK9jURyv5Iu7FslS5ZUTEyMKleurKZNm2rQoEG6+uqrgys0SAUtV7ja1dR8gKKKAAIU E/Pnz9fEiRNdPjt58qSOHj2qkiVLGqvDETw2bNhgbJ7RcurUKUlSXl6efv31V/34449av3693n33 Xc2dO1eVK1eOcoUojO+++05//PGHSpcu7dHv9OnT2rlzZxSqujQ49q3c3FwdPnxY27Zt05IlS/Tk k0+qa9euUa4OQGFxCRZQTGRkZPjsl5SUZKyOfv36XRLhw59vv/1W9913n86ePRvtUlAIJ06c8HmG Y+PGjcrNzTVc0aXt+PHjeuKJJ5SVlRXtUgAUEmdAEBW9e/eOdgnFzsGDB126Z8+erQ4dOigvL0+n T5+OUlWXrqysLC1cuFD33Xdf2KfN5RjmrF+/Xq1atfL6OczLzc3V1KlTNX/+/IhMv6jsW0WlDiBS CCAIWaAHyDfeeENjxoxxdqelpempp56KUFWXrry8PJfuTp06SZKuuOKKaJRTbOXf7s+dO6fff/9d +/bt04oVK/Taa6/p119/dfZfvnx5RAIIIisuLk7Hjx+XJP33v//1Okz+z/MPj9Dk36/sdrvsdrv+ +OMP5eTk6JlnntHnn3/u7P/tt99GoUIA4UQAQUStXr1azz77rLO7QYMGmjlzps8bdHfv3q05c+Zo /fr1OnLkiEqUKKGqVauqefPmGjRokGrXru11PPcb9jZv3qxhw4Zp48aNKl26tG6++WZNmjTJZRi7 3a6PP/5YS5Ys0datW3X06FFJUoUKFZScnKw777zT+SU+krzdbLh//37NmjVL//nPf3Tw4EGVKFFC V199tbp166YBAwa43NPh62bO/J+H8mva8uXL9f7772vnzp367bffZLfbFRMToyuvvFI33nijBg0a pJiYmELXEeo6d1i9erXeeOMNbd26Vbm5uYqLi1PTpk01ePBgNWjQIOjlDkbJkiUVGxurunXrqm7d urrmmmt0//33O/v7avfCbnuB3qD6xRdf6O2339a2bdt09OhRnT17VpUqVVLNmjXVokUL3XXXXUpI SPA5n8KOH4nlDGbfCNW1116rzz77TJKUnZ0tu90ui8Xi7H/+/HllZ2d7Hb4g4TruhGu7L+z+FwkW i0UWi0Vly5ZVgwYN9MILL6h58+bO/t7O6Aa6T5i6+buw6yeUOguzb6xcuVJvvPGGsrOz9fvvv6ti xYpq166dhg4dqoSEBG6KR9gRQBAx27dv14gRI5y/zFevXl2vvvqqYmNjvQ6fkZGhZ5991uO66hMn Tmj37t1atmyZRowYoXvvvbfAeT/22GNavXq1s7tWrVou/ffs2aPhw4d7/XUzNzdXP/74o1asWKGW LVtq2rRpfr9khduHH36oJ554wuWXdOnCJT1ZWVlauXKlMjIyVKpU5HbfESNGeH36T15eno4cOaJv v/1WH3zwgd5++21VqFAh5PkUZp2fP39ejz/+uN5++22Xz0+ePKnly5frs88+08iRI0OuLRRpaWku 3Y4bafMzte098cQTyszM9DqPPXv26LPPPtMrr7yioUOHuoSmcI0fieU0tW80a9ZMW7Zs0dGjR/XL L78oOztbKSkpzv7Z2dnO4FCxYkU1bdo0oAASjjYJ53YfzmNuJP3xxx8u3X/605+iVEnBonVcCnXf sNvteuyxxzzqzcvL05tvvqmVK1dqxowZYa8X4CZ0RMTBgwf1wAMPOP9IX3HFFZo1a5aqVavmdfgP PvhATz/9tN+bOnNzczVhwgS9++67Bc7fET5iY2PVvn17DRw40Nlv7969uvvuu31eWpHf+vXr9ec/ /1kHDhwocNhwSU9P9/gjkt+6deu0YMGCiM3/zTff9PvoUYedO3fqhRdeCHk+hV3nU6ZM8fijmV9e Xp7L2TcTPv30U5du97Btatt7/fXXvYYHdydOnNDTTz+tDz74IKzjR2o5Te0b5cqVU+PGjV1qzC9/ d+PGjVWuXLkCpxmuNgnXdh/uY264nTt3TsePH9eaNWv04IMPuvTr3Lmz8XoCFa3jUqj7xtSpU/3W e/jwYT388MNhqRHIjwCCsDt16pQGDRqkH3/8UdKFP+aTJ09Wamqq1+Hz8vI0YcIE56/FcXFxGjdu nDZv3qysrCxNmjTJ5T6Gf/zjHzpx4oTfGi6//HJ99NFHys7O1sKFC10unxg+fLjLH/Yrr7xSL7zw grZs2aKsrCy9+OKLuvLKK5399+7dq0ceeST4hghRXl6eunXrpk8++UQ7d+7U66+/7hHcli9f7vx/ m83m9XS44/NgT5W//vrrzv9PTk7WsmXLtGvXLq1du1Y333yzy7Bff/11SHUUdp3v2bPHpU5Jat26 tZYtW6adO3dq+fLlatu2rdczEOHk+JK0fft2TZ06VaNGjXLpX6dOHZduU9veW2+95dL997//XRs3 btT333+vZcuWqX79+i79X3755bCOH6nlDHbfCFWZMmXUunVrZ7f7U93yd7dp00ZlypQpcJrhaJNw bfeROOYWVmJiosu/2rVrq2HDhurfv7/LU6+Sk5OL7BfiaB6XQtk39u3bp1dffdXls3bt2unf//63 du7cqUWLFqlp06Y6cuRI2OsFCCAIK7vdriFDhrj8wXj00UfVsWNHn+O8/fbbOnz4sLP7b3/7m+69 915dfvnlKl++vHr37u1yyvrIkSMFvtyuV69eqlevnsfnH3/8sctjNatUqaK3335bd911l+Li4lS+ fHl169ZNb7/9tss7HL7++mutWLHC/8KHyS233KIXX3xRderUUZkyZXT99dcrPT3dZZi9e/dGbP4v vviiHnroISUlJempp55SamqqLrvsMl111VX6v//7P5dh86+3YBR2nb/55ps6efKks7thw4ZauHCh UlNTVaZMGSUnJ+v111/3GXoLw9uXpFtvvVXTpk3zuBG5R48ezv83ue3t27fPpfvOO+9UpUqVVLJk SaWmpuof//iHatSooTvuuENPP/20Zs2aFbbxI7mcpvaNMmXKuFxOt2XLFpf++Y9vaWlpBQaQcLVJ uLb7SBxzTUhJSdGbb74ZUOCLhmgel0LZN958802Xh5ekpqbqtdde0zXXXKMyZcqoSZMmyszMNPoY d1w6CCAIq3HjxrlchtK/f/8CnwK0du1al+5bb73VYxj3U+75n4jizXXXXef183/9618u3f3793f5 1dHhyiuv1IABA1w+W7Rokd95hkufPn08PnNfnkj+st+gQQONGjVKK1euVJMmTVz6ubfV77//HtI8 CrvO3X+RHjRokMe1zaVKlVL//v1Dqi8c2rZtq+7duzu7TW577tfI9+/fXx999JHsdrskqVGjRvr8 8881ffp0/eUvf/G4wbQw40dyOU3tG6VLl1ZSUpJq1Kgh6UIg27Nnj6QLZ/QcAa1GjRqqU6eO1xcV 5heuNgnXdh+JY64JW7du1YABA5xn14uaaB6XQtk33Ou99957PeotW7asxzYJhAM3oSNs5s+fr4UL Fzq7b7zxRo0fP77A8Xbv3u3S3bZt2wLHycnJ8du/evXqXj93f4GVvzMzHTt21OTJk53d+Z964+uJ T/mF+pQQb782VaxY0aXb/ZG7kbJ37159+OGH2rdvn/bt26ddu3a59M//a18wCrvO3b+AtGzZ0us4 +Z+cY9JNN92kadOmuXwWrm0vEH379nV59HVOTo7++te/KjExUbfeeqt69uzp8WCGcI0fyeU0tW+U LVtW0oXtx7GtrVu3TldffbXWrVvnHK5FixYuw/sSrjYJ13YfiWOuKZs2bVKfPn20bNkyxcXFRbsc F9E8LoWybzhCtYNje3aX/3JEIFw4A4KwWLlypaZMmeLsTk1N1csvv+xy74UvoVxb7O9mO+nCPSDe HDt2zKX7qquu8jkN934FzTNcvD1VyvQlB1999ZXuvfdederUSc8++6zeeOMNrVq1Kmy/PBZ2nbsH n/j4eK/jVK1aNej5BCMmJkbly5dXlSpVVL9+fd1xxx16/fXXNW/ePI8b0E1ue3/5y1/UtGlTj89t NpteeeUVdejQQffcc4/Ll+lwjR/J5TS1bzimmf+L+TfffOPy3/z9C6ohXG0Sru0+Esfcwsp/r5jN ZtPu3bu1detWffrppxo7dqzLsu7Zs0czZ84Meh7nz58PZ8keonlcCmXfcA8kvury9fAYoDA4A4JC y8rKUnp6uvPgW6NGDc2bN8/lHRH+nDt3Luh5uj+W0Z2vebu/f8RfQHJcbhLIsOF02WWXeXxmat52 u13p6eku13vHx8fr+uuvV8eOHdWyZUuXpwOFqrDr3L09zp496/UZ977eN1MYoZ7ZMrntlShRQnPn ztWbb76pDz/8UNu2bfMY5quvvtK3336rXr16eZypLMz4kVxOU/uG44xGWlqaYmJidOrUKW3evFnS /+4HiYmJcd4nUtAZkHC1Sbi2+0gcc8OtVKlSslqtslqtql27tmrUqOHyuOeVK1dq9OjRPsd3f3eL FPlliOZxKZR9w31bc+8OdDpAKAggKJT9+/frwQcfdP7CV7FiRc2ZMyeoX3isVqvLLzHZ2dk+3xUS KG8HY+nCr0T5f/3bu3evzxvs3H/tr1SpUqFqClQ0D/YzZsxwCR9du3bVs88+6wx0oXxx8aaw69xq tbqsx59//llXX321x3A///xz4QoNI9PbXsWKFTVkyBANGTJEu3fv1gcffKBPP/1Uu3fvdl4LfurU Kb3++uuKj4/XkCFDwjJ+JJfT1L7huKejUqVKqlWrlrKzs/XTTz9p165d+umnnyRJtWvXdl7iUtA9 IOFqk3Bt95E45kZau3btXLoPHjzod/jff//d44eoSL+tPprHpVD2jdjYWJfHMP/8889ez87t37+/ ULUB3nAJFkKWl5enQYMGOW/IjI2N1fPPP69rrrkmqOm430/h/sSZUPg6GLs/PnTlypU+p/HJJ5+4 dDds2ND5/+6XC3j7dzFyf//HxIkTXf6I//LLL2GZT2HXec2aNV26818Wk99XX30V1HQjKVzbXijq 1KmjYcOGadmyZfrwww/Vs2dPl/4FvTsjmPGjuZzhkv+MhuNG3lOnTmnmzJnO8JX/+v6CzoCEq03C td1H4pgbaVu3bnXpdr9ZuqBLHiXPex7C7WI7LrlvB/kfq55fUXgAAYofAghCcv78eT388MMul2aM Hj1aN9xwQ9DTat++vUv33LlzC12fL7fccotL9+uvv+711539+/d7PM89/yNVi6tDhw65dP/2228u 3d5eUOh+2t79i8CZM2c8xinsOm/VqpVL9/z58z0urzh9+rTmz58f1HQjydS2Z7PZlJmZqaFDh6pd u3YeL76rWbOmJkyY4PLZ2bNnwzZ+cdjH8l87n/+X9/zBIf82XNC19uFqk3Bt9yaPuYVx9uxZ/fLL L1q8eLHHO1Hcv+xbrVaXbm9f/jMyMsJfZD4X23HJvd4FCxa47MvShZdRFpV6UbwQQBCSJ554QmvW rHH5bMyYMR4vk/L3z6FXr14uT+tYtWqVHnnkEf3www86e/as9uzZozFjxqhdu3YaOnSoMjMzQz7D cMcdd7i8H+TQoUPq1auXFi9erBMnTig3N1dLlixRr169XL6Mp6WlubwXoLhyf6rM6NGjtX//fh0/ flyzZ8/WjBkzPMZxv/HS/bKH//znPzp06JC2b9/u/Kyw6/yee+5xqXXbtm0aMGCAtmzZotOnTysr K0v9+vXTzp07Q2qHSDC17Q0ePFhPPPGEli5dqh9//FHDhw/XypUrlZubK7vdrhMnTni8aDD/E60K O35x2Mfyn9Fo1aqVc1tzXK4SFxfn8ojTgs6AhKtNwrXdmzzmBsrb34g6deqoadOmGjZsmMe7abp0 6eLS7R5I/vGPf+jzzz/X6dOnlZOTo7///e9asmRJRJfhYjsu9erVy6XerKws3Xfffdq5c6dOnz6t 9evX6+67775oz+ijaLPYbDbvdx0BfgTyGNqC5D+oLViwQOPGjQt43Jtvvllz5szxWY+/A+b27dvV t2/fgN/umpSUpHfeecfjkYaBKqi2QGs3MZ3HHnvM48tlQT755BOXN3736NHD4/ny0oVfgfO/sK6w 6/zVV1/Vk08+6Xec2NhYjye9BPvHNJhtqyDh2vb81fTFF1/or3/9a1CPo33hhRd01113hWV8ycxy hjKcL+7jb9y40eXei969e+vLL790drdp00aZmZnO7iNHjqhZs2Z+awhXm4Rruzd5zPWmMH9DkpKS 9MEHH7iceXrjjTdcHh3tjdVqdbnnQQr+OFpQ/3CtH1N/N6ZOnerxyHB3tWrV8ngMM6EEhcUZEBQJ /fv316hRowK6EfLmm28u8IDpzzXXXKO33npLjRo1KnDY9u3b6+233w45fFxsRo0apbp16/rs37Bh Qz300EMun7k/ivWee+7xOu7333/v0l3YdT5w4ECXp+K4K1eunN+n5ESDiW2vbdu2mjx5ckA3rsfE xOihhx5yCQ+FHV+6+Pcx9zMa7u9BcO8u6AyIFL42Cdd2b/KYG061atXSnDlzPC57+8tf/qLrr7/e 53hWq1Vjx46NcHUX33Fp2LBhuvPOO332r1u3rsv7vSTfT5kEgsEZEIQk3GdAHHbv3q0FCxbo66+/ 1s8//6w//vhDZcqUUZUqVZSamqru3bt7fSlSKL/G2e12ffLJJ1qyZImysrL0yy+/yG63q2LFikpN TVWPHj3UoUOHEJcu8NqK0hkQ6cJlJtOnT9enn36qgwcPqkSJEkpMTNSdd96pfv36KScnx+VFao0b N9bixYtdpvHuu+9qwYIF2rNnj+x2u+Lj49W0aVM9//zzHvWEus4d1q5dq3nz5mnr1q06ffq0Klas qEaNGmnAgAFq1qxZ2H+pDccvf4Xd9gKp6dixY8rMzNQXX3whm82mEydOyG63q2zZsqpWrZoaN26s e+65R8nJyV7nUdjxTS1nMMMFOp/vv//e5fGpWVlZuv32253dy5YtU2pqqrP77NmzLmcB/dUQruNO uLZ7k8dcf+PnFxMToxIlSqhEiRK67LLLFBMToyuvvFI33HCD+vfv7/ML8Pnz5/Xaa69p0aJF2rNn j86fP68qVaqoWbNmGjRokOrVq1fo42igy13Y9WPqeO+wePFiZWRkaPfu3Tpz5oz+9Kc/6bbbbtOD Dz6osmXLukzn8ssvdz6WGggVAQQAAABenT592uUepho1avBkLBQa7wEBAAC4xNjtdqWlpalx48Zq 2bKl2rRp4/XM1KZNm1y6a9SoYahCFGcEEAAAgEuMxWKRxWLR0qVLtXTpUkkX3v3UrVs3lS9fXseO HdP69ev1zDPPuIzXtm3baJSLYoZLsAAAAC5B06ZN09SpUwMevmrVqvr44491+eWXR7AqXAp4ChYA AMAlaMiQIS4PWPAnISFBL7/8MuEDYcEZEAAAgEvYqlWrtGTJEm3fvl2HDx/W77//LovFIqvVqquu ukrt27dX//79CR8IGwIIAAAAAGO4BAsAAACAMQQQIEI+/PBD3X777UpJSVHDhg116623as6cObLb o3PSMTExMagXSAY7vCR9/vnnuv3225WamqrGjRurd+/ewRUZYe7LFMoyhjqvSCloPqbqCEYgNYWr 7miu8+KwfUVKOOq/2NsAuJQRQIAImDdvnh5++GFlZWUpNzdXx48f1/bt2zVp0iQ999xzzuHWrl2r Hj16RLHS8Bo9erSysrJ04sQJHTt2TOXKlYt2SbgIFOX9oCjUVhRqKC5oS6Bo4D0gQATMmTNHkvTy yy/rxhtv1Llz5/Tee+9p/Pjxev/99zVq1ChJUr9+/aJZZtjt27dPkrRmzRpVr15dv//+e5Qrgs1m i3YJHtxrMrkfBNsewdQWqbb2VkNRXK8Xg+J2zAUuVpwBASLg3LlzkqQdO3bozJkzslqtGjBggGw2 m77++mtJ8nupxjvvvKMOHTooJSVFHTt21AcffOAyfcfw8+fP13XXXadmzZpp7ty5QdUY6iVZy5cv V8eOHZWSkqLbb79dmzdv9lietLQ01alTR1arVZK0ePFi3XjjjUpOTva7PAsXLtS1116r7t27u3z+ 6quvqnnz5kpLS9P27du1YMECNW3aVGlpafriiy+c09m1a5d69uyphg0bKiUlRZ07d9Y333wT0PK1 a9dOiYmJ+uGHHyRJOTk5SkxMVPv27X2Os2DBArVu3VqNGjXS8OHD9ccff3gMk5OTo27duiklJUV3 3XWXdu3a5dK/oHUd6Hwk6fDhw0pLS1PTpk2d8/F1WZCv9egwe/ZsNW/eXNdff73Wr1/vd3vp0KGD EhMTtW3bNknSSy+9pMTERD3//POSpJ07dyoxMVEdOnTwqKmgS5YKaj93M2fOVIsWLXTddddp4cKF Hv3d57Flyxbn9Bs2bKgePXrou+++81tbQduruwULFqhly5Zq0aKF88cJf8scSPt4Gy/Q/aygdZ9f QftUoNMsaL24C/UY568NfLWlv20AQGTwFCwgAsaNG6cFCxZIkq644gp16dJF999/v66++mrnMN6+ qNhsNi1dulRDhw51+bxcuXKaN2+eWrdu7TJuTEyMTp065Rzu+eefd34ZcucYx/HLqXt3oMO7zzM1 NVXLli3zuTwrV67U0KFDXcaJiYnRjBkzXL6Q5p92mzZtlJmZ6XWaSUlJ+umnn5zTS0pK0sqVKyVJ t912m7Kzs12Gr1Onjj755JMC2+CRRx7RkiVLNH78eA0YMECvvfaaJkyYoLvuuksvvPCCRx2LFi3S 8OHDXT4bPHiw0tPTXaZdq1Yt5eTkOIdp1qyZ3nvvPUkKaF0HOp/s7GzdfffdysnJ0cyZM5WWluZ3 mX2tR0n65z//qZEjRzr7Va9e3Xl2y9v2MnHiRM2fP1+PPvqohgwZooEDB+rTTz9VWlqaFixYoJkz Z+q5557Tfffdp7Fjx7rU5Gu7CaT93GVmZuqJJ57w2s/Xdt+xY0ePUNOgQQN9+OGHBdbma3t1n5e7 SZMmOe+P8rYfBtM+jvFC2c8c8q97d4HuU/6mGch6cRfIMS7YNvDVlv62AQCRwRkQIALGjBmjgQMH qnz58jp27JgyMjLUuXNnpaenKy8vT5LrH16bzebsnj9/vqQLf2h37typ6dOn6+TJk5o3b57HfG65 5RZt2bLF+Qf5tddeC7jG/PMMxi233KKtW7c672VxnDHwtTyzZs3SqVOn1LVrV23atEndunXTqVOn NGPGDI9pd+3aVdu2bdPTTz/t8vn48eO1ePFiSRd+kR05cqTeffddSdL+/fudwy1fvlw2m00//PCD Fi1aJEk6cOBAQMvVqlUrSdKXX34pSfrqq69cPnfnWE9Tpkxx1uLtS1yTJk2UnZ2tadOmSbpwNsB9 Gv7WdaDzefDBB5WVlaWRI0c6w4c/vtajJOcv1N27d1d2draaN2/ud1qdOnWSJOfZqG3btikmJkbb t2+XdOHhBPmHy8/XduPgr/3cZWRkOOvOysrSXXfd5bdu6X/bx9KlS53zd3zxLKg2X9uru549eyor K8u5nzrqDERBNTgEs5/5W/fuAt2n/E0zlPWSf7qBHuMKagNfbelvGwAQGQQQIAJKlSqlMWPGaNWq VRo8eLAqVKigvLw8vfvuu3rsscf8juv4tXfEiBGqV6+e8xdyx+Ut+Q0dOlRxcXF6+OGHJUl79uwJ 85J4Gj16tKxWq7p27SpJys3N9Tv87t27JV0IERUqVNC4ceMkyeulNPfff7/KlSvn8Uul4xKQ/N1N mzaVJGegc9iwYYPGjBnjvM/Gvb8v7dq1kyR9++23OnfunDZt2iRJPi/Bcnx5ueOOO9S8eXPZbDat XbvWY7j09HTFxsbqtttukySdOHHC2S+QdR3ofBxf8q+66qqAltffenRsR4MHD1ZsbKwGDx7sd1ot WrRQQkKCtm7dqh07dujAgQO69tprdejQIW3fvl1ZWVlKSEhQixYtAqotP3/t5+7HH3+UdGG/KF++ vB566KECp+84O3D33Xfr9ttv17PPPqtff/01oNp8ba/uHPU8+OCDkqS9e/f6HPb8+fMBzdtdMPtZ sPtwIPuUv2mGsl4cgjnGBdMG+RVmGwAQGgIIEEGVK1dWenq6Pv74Yz3wwAOS5PXLY36+voB4++Ll uNTAMY7FYilMuQGpUqWKJKlMmTIBDe9ek7/HEF955ZVeP4+NjVWJEv87XFmtVpduh1mzZqlv374q WbKkJk+eHFB9DtWrV1etWrV06NAhvfXWWzp8+LDq1q2rqlWr+h2voC+MjvFLlfJ85kcw67qg+TRu 3FiVK1fWyy+/7Hc4B3/r0bGOHPcyxcTE+J2WxWJR27ZtlZub65z/fffdJ0maMWOGcnNz1aZNm5C2 T3/t562O/PUGMr8pU6booYceUv369XX48GHNmjVLgwYNCqg2X9urO8fbo0uWLClJXrddx309R44c CWia7oLZz4LZhwPdp/xNM5T14hDMMS6YNsivMNsAgNAQQIAIaNWqlRITEzV79mydOHFC5cuXV506 dSRJZcuW9Rj+6NGjzl/catasKUmaPHmyvvvuO+clAe7XYUsXvtzl5eVp9uzZkgY4wBoAACAASURB VC5cLx9pwX6JdCzPhAkTdOzYMU2cOFGSVK9ePY9hAw01vjguwejdu7fLJSCBfhG59tprndORLtxv 4IvjV+9ly5Y5b9L29gu/v/YKZF0HOp+5c+eqS5cu2rhxo9asWeN/QQuoq3r16pIutENeXp6mT59e 4PQcl1etWrVKlStX1o033qiEhAStWrVKknTzzTcXOI38+0EgdbpztNVTTz0VcN2lS5fWqFGj9K9/ /ct5qdCOHTsCqi3Q7dVxad0rr7wi6X/rXZLi4uIkXXhvUF5enqZMmeJzOt5qcAhmPwumTQPdp/xN M5T14hDMMS6YNsjfloFuAwDChwACREC3bt0kSc8884xSU1NVv3595029999/v3O4SpUqSbrwxXfE iBGSpIEDB0q6cOlJ3bp1nU9r6du3r8d8Vq1apeTkZOdNuffee2/ANXp7ik4kOJZn8eLFaty4sRYv XqyYmJgCL+sJRfny5SVd+DI8bNgwVahQQZJ08ODBgMZ33Pj9008/SZLatGnjc9j+/ftLurCeevXq Jel/l3IEKpB1Heh8KleurH79+ikmJkYvvfRSUHW4u+eeeyRJ77//vpKTk/0+JckhLS1NFSpU0MmT J3XNNddIklJSUnTy5EldccUVuv76632O620/CIXjrMvSpUuVnJzsfDiBP126dHG2+0033SRJql+/ flhry8zMVIMGDZz38OTfT2vXri1JGjZsmJKTk70+BS2QGiK1nxV2n5JCWy8OwRzjAmkDb21Z0DYA IPwIIEAEpKen6+9//7uSkpJktVp1+eWXq3Hjxnr55ZedfyQlaciQIapSpYqsVqvzsptu3brp2Wef VVJSkmJjYxUfH6+uXbtq5syZHvMZO3asqlatqipVqmjixInO66+Lki5dumjq1KmqV6+erFarkpKS NGXKFL9fSEP1zDPPqHbt2rJarWrdurXzZu4XX3wxoPHbtm3rvEwkNjZWbdu29Tlsjx49NGbMGCUk JOiKK67QHXfcoTFjxgRVbyDrOpj51KxZU82aNdOGDRsKvNTPnwEDBujFF19U9erVVa9ePecN4P5e LFm6dGnnmRnH/TqOM0rNmzf3e7bA234QijvvvFNjx45VfHy8EhISXJ7k5ctLL72kNm3aKC4uTuXL l1erVq1cnnoWjtpGjRqlhIQExcfHa8yYMc4fKKQLT8RKTU2V1WpVrVq1NGnSJI/xA6khUvtZYfcp KbT14hDMMS6QNvDWlgVtAwDCj8fwAhehgh6hi9B17txZ27ZtU8uWLfXOO+9Euxzj7Ha7kpOTdfLk SWVkZKhFixZaunSpHn30USUkJGjdunXRLhGXAI5xQPHGm9AB4P87fvy48+3tgdyzUBxZLBY1btxY X375pfr06ePSz/GkMAAACoNLsADg/2vdurV+++03de/e3ePL96Vk6tSp6tSpk+Lj41WuXDlVrVpV 3bt319ixY6NdGgCgGOASLAAAAADGcAYEAAAAgDHcAwLAJ/fH9MbGxqp8+fLq1auXhg8f7jGc6RtG Tc23oPmYquPzzz/Xc889J5vNppIlS6pBgwZ68803IzpPbga+OAW73gIZnm0BQLhwBgRAwPLy8vTz zz9r2rRpmjp1aqGmtXbtWvXo0SNMlV0aRo8eraysLJ04cULHjh3z+1jcUFyM6+RirBkALnWcAQFQ IJvNJrvdrt9++02zZs3SrFmz9N5777mcBQlWv379wlhhdJn6RXjfvn2SpDVr1qh69erOJ3aFy8W4 Ti7Gmk3gLAWAoowzIAACYrFYdMUVV2jo0KGSpGPHjvkdfvHixbrxxhuVnJysjh07urzhOf+lXQW9 kX3BggVq3bq1GjVqpOHDh+uPP/7wGCYnJ0fdunVTSkqK7rrrLu3atcul/zvvvKMOHTooJSXFo5Zg 5iNJhw8fVlpampo2beqcj/syOLqXL1+ujh07KiUlRbfffrvHG8Vnz56t5s2b6/rrr9f69ev9tkX+ z9PS0lSnTh1ZrVZJ/ts6fz0LFy7Utddeq+7du/udvrc6wtHG7gKt29dn/mouaH0GOu9XX31VzZs3 V1pamrZv364FCxaoadOmSktL0xdffBFyGyxfvlyJiYkeb/YeMGCAEhMTnePu2rVLPXv2VMOGDZWS kqLOnTvrm2++8ajTfd26t0dB08nfbi1btlSLFi00Z84cn/WHsswA4MBTsAD4lP+ab7vdrmPHjmnm zJmaO3euEhMT9dlnn3kMJ0krV67U0KFDderUKee0YmJiNGPGDHXo0MHrl2xvv9guWrTI4yzL4MGD lZ6e7jLfWrVqKScnxzlMs2bN9N5770mSli5d6gxNDuXKldO8efPUunXroOaTnZ2tu+++Wzk5OZo5 c6bS0tK8Lr+jOyYmxqUNUlNTtWzZMknSP//5T5c3QlevXt15hsNbW/hqs4La2ls9bdq0UWZmZkDT D1cbuwum7vztkf8zXzUXtD6DmXd+SUlJ+umnn5zjJSUlaeXKlSG1wdmzZ9W2bVudOHFCq1evVpUq VXTw4EF16NBBVqtV//nPf1SqVCnddtttys7Odhm3Tp06+uSTT1zqdF+37m0X6HTcTZo0Sb1793YZ xjHNUNY7AEicAQEQgMTERNWsWVNNmjTR3LlzJUm9evXyOfysWbN06tQpde3aVZs2bVK3bt106tQp zZgxQ5LrF0qbzebzcpH58+dLkqZMmaJ3331Xkpxf4PNr0qSJsrOzNW3aNEnSzp07Pabx/PPPa+fO nZo+fbpOnjypefPmBT2fBx98UFlZWRo5cqQzfPhzyy23aOvWrXruueckST/88IOz38KFCyVJ3bt3 V3Z2tpo3b+53Wr7arKC2zq9r167atm2bnn766YCn71DYNnYXTN2++Kq5oPUZzLzHjx+vxYsXS7pw FmHkyJHOae7fvz/kNnCEi7y8POf03n33XeXl5emWW25RqVIXrpBevny5bDabfvjhBy1atEiSdODA AY/p+Vu3wUynZ8+eysrKcp5JycjI8Dq9UJYZABw4AwLAJ/dfRcuXL6/KlSurZ8+eeuihhzyGc3wB bNSokX777Tdt2rRJFSpU0K+//qomTZooLi5OW7Zs8TqON6mpqTpx4oS+++47lS5d2md969atU0JC gs6ePas6deq4TLdhw4Y6fvy4x7gJCQlat25dUPNxeO2113TDDTf4XH5H9/r161W1alWdPn1a9erV cxnGMc9Vq1apVq1a2rVrlzp27Oi3Tby1WTBt/emnn6p27dpep+1r+uFqY3ehbCPnz59XrVq1XD7z VnNB6zOYeWdnZysmJsY5361bt6pcuXIedYTSBo51XrduXX388cfq2LGjdu3apRUrVqh+/frO4TZs 2KDFixfr66+/1nfffed1+d3Xrbd2CWQ6n3/+uWrUqOGszWq1auvWrV6nGcoyA4DEGRAAAXD8upyV laVVq1a5hA9vLBaLS7fdXrjfOc6fP++3f9WqVSXJ+atxIOOeOHEi6Pk0btxYlStX1ssvv+x3OIcq VapIksqUKePRz9Em586dk3ThEppQBNPWV155ZUjzkMLXxg7B1O24f+PIkSMF1hlIXcHMOzY2ViVK /O9PpdVqdekuaF7+2iApKUnNmjXTd999p7feeku7du1So0aNXMLHrFmz1LdvX5UsWVKTJ0/2Oa2C 1m2g07n88sslSSVLlpQkr8vqEMoyA4BEAAEQATVr1pQkTZgwQceOHdPEiRMlyXkWIL+jR4/q119/ 9Todxy+uy5Ytc96k3aJFC4/h3L9Qeqtl8uTJ+u6775xhKv/18IHOZ+7cuerSpYs2btyoNWvW+Jxn IHVVr15d0oUvhnl5eZo+fXqB0/MmmLb2FoS88bZOCtvGodQdFxcnSfrwww+Vl5enKVOmBFRzQesz mDYLVChtIMl5qZNj2e68806X/o7LxXr37u1yCZ97aCpo3QY6HcdlVK+88orLcnkT6jIDAAEEQNgN HDhQ0oUnDTVu3FiLFy9WTEyMBg8e7BymUqVKkqRrr71WI0aM8Dqd/v37S5LS09Od95w4bhIOtpb0 9HTVrVvX+XSgvn37Bj2fypUrq1+/foqJidFLL70UVB3u7rnnHknS+++/r+TkZI8nZAUqkLYOVCDr xF8N/to4lLodlxQNGzZMycnJXp+w5K3mgtZnONvMfZrBtIEkdevWTRUrVtQvv/yiuLg4jyeUlS9f XpLUqVMnDRs2TBUqVJAkHTx4MKj6Ap1OZmamGjRo4Lwvxf0pXfmFuswAQAABEHZdunTR1KlTVa9e PVmtViUlJWnKlCm6/vrrncMMGTJEVapUkdVq9XnJRo8ePTRmzBglJCToiiuu0B133KExY8YEVUu3 bt307LPPKikpSbGxsYqPj1fXrl01c+bMkOZTs2ZNNWvWTBs2bNDatWuDqiW/AQMG6MUXX1T16tVV r149583dwb5cMJC2DlQg68SbQNo4lLonTZqk1NRUWa1W1apVS5MmTQqo5oLWZzjbrDBtIF249O6m m26SdOHxyo6g4PDMM8+odu3aslqtat26tfMG7xdffDGo+gKdzqhRo5SQkKD4+HiNGTNG3bp1C/sy AwA3oQOAYXa7XcnJyTp58qQyMjLUokULLV26VI8++ig38AIAij3ehA4AhlksFjVu3Fhffvml+vTp 49KvXbt2UaoKAAAzuAQLAKJg6tSp6tSpk+Lj41WuXDlVrVpV3bt319ixY6NdGgAAEcUlWAAAAACM 4QwIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAA wBgCCAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAA AMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCkV7QKiLTExMdol AAAAAMWGzWbz2/+SDyCSZLfbo10CAAAAcNGzWCwFDsMlWAAAAACMiXoAOXr0qFq3bu3x+ebNm9Wh QwelpKSoc+fOWrduXUD9AAAAABRdUQ0gP/74o/r06aP9+/d79Js9e7a6du2qDRs2qF27dpo/f35A /QAAAAAUXVENID179tSf//xnr/127Nih3r17q2zZsurdu7dycnIC6udNYmKiz38AAAAAzInqTeiL Fi1StWrVNH78eI9+R48eVYUKFSRJCQkJOnToUED9vPF3Jz4hBAAAADAnqmdAqlWr5rPf6dOnVarU hXxUunRpnTlzJqB+AAAAAIquqN+E7stll13mDBZnzpxRmTJlAuoHAAAAoOgqsgEkPj7eeWnVoUOH VLFixYD6AQAAACi6imwASU5OVkZGhnJzc5WZman69esH1A8AAABA0VVkA0ifPn20YsUKtWzZUitW rNADDzwQUD8AAAAARZfFZrPZo11ENCUmJspuv6SbAAAAAAgLi8Xi9wm0UhE+AwIAAACg+CGAAAAA ADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAA AAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAA AAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCGA AAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwh gAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACM IYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAA jCGAAAAAADCmVLQLKIosFktI49nt9jBXAgAAABQvBBAfPtp0IKjhOzWpFqFKAAAAgOKDS7AAAAAA GEMAAQAAAGAMAQQAAACAMQQQAAAAAMYQQAAAAAAYQwABAAAAYAwBBAAAAIAxBBAAAAAAxhBAAAAA ABhDAAEAAABgDAEEAAAAgDEEEAAAAADGEEAAAAAAGEMAAQAAAGAMAQQAAACAMQQQAAAAAMYQQAAA AAAYQwABAAAAYAwBBAAAAIAxBBAAAAAAxhBAAAAAABhDAAEAAABgDAEEAAAAgDEEEAAAAADGEEAA AAAAGEMAAQAAAGAMAQQAAACAMUU2gOzYsUNdunRRcnKyunbtqu+//97Zb/PmzerQoYNSUlLUuXNn rVu3LoqVAgAAAAhUkQ0g6enpSk9P15YtW/S3v/1N6enpzn6zZ89W165dtWHDBrVr107z58+PYqUA AAAAAlUq2gX4kpOTozZt2qhkyZJKS0vTsGHDnP127NihCRMmqGzZsurdu7cGDhzod1qJiYkRrhYA AABAIIpsAKlfv76+/PJLtW3bVuvXr1fdunWd/Y4ePaoKFSpIkhISEnTo0CG/07LZbD77EU4AAAAA c4psAHn88cfVv39/5ebmymq1asGCBc5+p0+fVqlSF0ovXbq0zpw5E60yAQAAAAShyN4DMm7cOM2Z M0fff/+9pk+frieffNLZ77LLLnOGjjNnzqhMmTLRKhMAAABAEIpsANmzZ49atWqlkiVLql27dtq9 e7ezX3x8vPOyq0OHDqlixYrRKhMAAABAEIpsAElKStI333yjs2fP6uuvv9bVV1/t7JecnKyMjAzl 5uYqMzNT9evXj2KlAAAAAAJVZO8Beeqpp5Seni6bzabExERNnjzZ2a9Pnz4aPXq0Fi5cqCpVqmjq 1KlRrBQAAABAoCw2m80e7SKiKTExUXa7axNYLBZ9tOlAUNPp1KSax3QAAACAS4nFYvH7BFqpCF+C BQAAAKD4IYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABj CCAAAAAAjCGAAAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAA YwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADAGAIIAAAA AGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAA AABjCCAAAAAAjCGAAAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggA AAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADAGAII AAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgC CAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAY AggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADA GAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAA wJgiG0DOnTunkSNHqlGjRmrbtq3Wrl3r7Ld582Z16NBBKSkp6ty5s9atWxfFSgEAAAAEqsgGkFdf fVVxcXFat26dRo8erbFjxzr7zZ49W127dtWGDRvUrl07zZ8/P4qVAgAAAAhUqWgX4Mu///1vjRs3 TjExMerSpYu6dOni7Ldjxw5NmDBBZcuWVe/evTVw4EC/00pMTIxwtQAAAAACUWQDyN69e7V69Wrd d999qly5smbOnKnatWtLko4ePaoKFSpIkhISEnTo0CG/07LZbD77EU4AAAAAc4rsJVinTp1Sbm6u vvrqKz3yyCMaM2aMs9/p06dVqtSF7FS6dGmdOXMmWmUCAAAACEKRDSAlS5bU0KFDVaZMGd10003a unWrs99ll13mDB1nzpxRmTJlolUmAAAAgCAU2QBSuXJlnT59WpJ0/vx55xkPSYqPj3dednXo0CFV rFgxKjUCAAAACE6RDSDt27fXW2+9pXPnzmnlypVq0aKFs19ycrIyMjKUm5urzMxM1a9fP4qVAgAA AAhUkQ0gjz76qDZv3qxGjRrp1VdfdXkMb58+fbRixQq1bNlSK1as0AMPPBDFSgEAAAAEymKz2ezR LiKaEhMTZbe7NoHFYtFHmw4ENZ1OTap5TAcAAAC4lFgsFr9PoJWK8BkQAAAAAMUPAQQAAACAMQQQ AAAAAMYQQAAAAAAYQwABAAAAYAwBBAAAAIAxBBAAAAAAxhBAAAAAABhDAAEAAABgDAEEAAAAgDEE EAAAAADGEEAAAAAAGEMAAQAAAGAMAQQAAACAMQQQAAAAAMYQQAAAAAAYE3IASUxM9Pjs+PHj6tmz Z2HqAQAAAFCMlQp2hB49emjDhg2SvIeQyy+/vNBFAQAAACiegj4DMnToUJ/9KlWqpMcff7xQBQEA AAAovoI+A9K+fXvZbDYlJibKZrNFoCQAAAAAxVXI94AQPgAAAAAEK+gzIA6LFi3S888/r19++UWn Tp1y6Uc4AQAAAOBNyAHkmWee0aFDh8JZCwAAAIBiLuQAUqpUKWVkZOi6665TqVIhTwYAAADAJSTk e0CefPJJffPNN/rtt9/CWQ8AAACAYizkUxcDBw6UJE2bNs2jH/eAAAAAAPAm5DMgAAAAABCskM+A cJYDAAAAQLA4AwIAAADAmJDPgCQmJvrsx9kRAAAAAN5wBgQAAACAMWG5B8Rut+vXX3/VtGnT1KFD h3DUBQAAAKAYCssZEIvFori4ON1yyy0aP358OCYJAAAAoBgK+z0gCQkJoU4SAAAAQDEXcgBxV65c OVWpUkUjR44M1yQBAAAAFDO8BwQAAACAMTwFCwAAAIAxIQeQc+fOacqUKWrXrp2Sk5PVvn17TZs2 TefPnw9nfQAAAACKkZAvwRo/frzeeOMNZ3deXp6mTp2q3377TWPHjg1LcQAAAACKl5DPgCxfvlzj xo3Tf//7X+Xk5Gjjxo0aO3aslixZEs76AAAAABQjIQeQEiVKyG63y2KxOP85ugEAAADAm5Avwerc ubMmTpyoiRMnunzer1+/QhcFAAAAoHgKOYCMHTtWVqtVy5Yt06+//qoKFSro1ltvVXp6ejjrAwAA AFCMWGw2mz3aRURTYmKi7HbXJrBYLPpo04GgptOpSTWP6QAAAACXEovFUuD7AoO6ByQ3N1d9+/b1 2b9fv37Ky8sLZpIAAAAALiFBBZDnnntOn3/+uc/+a9eu1ZQpUwpdFAAAAIDiKagAsnr1ao0cOdJn /xEjRmjVqlWFLgoAAABA8RRUADl69KgGDBjgs/+9996rI0eOFLYmAAAAAMVUUAEkNjbW700l+/bt U0xMTGFrAgAAAFBMBRVA2rZtq4ceekjvv/++Dhw4oHPnzuns2bPat2+f3nvvPT3wwANq1apVpGoF AAAAcJEL6j0go0ePVq9evTRixAiv/WvUqKHHH388LIUBAAAAKH6COgNSpUoVLV68WPfff7+SkpIU Fxcnq9WqWrVqqX///lqyZImqVasWqVoBAAAAXOR4ESEvIgQAAADCIuwvIgQAAACAwiCAAAAAADCG AAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAw hgACAAAAwBgCCAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAA MIYAAgAAAMAYAggAAAAAYwggAAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCGAAAAA ADCmyAeQtWvXKjEx0eWzzZs3q0OHDkpJSVHnzp21bt266BQHAAAAIChFPoC89NJLHp/Nnj1bXbt2 1YYNG9SuXTvNnz8/CpUBAAAACFapaBfgz9q1a2WxWDw+37FjhyZMmKCyZcuqd+/eGjhwoN/puJ9B AQAAABAdRTqAvPTSS3rkkUfUp08fl8+PHj2qChUqSJISEhJ06NAhv9Ox2Ww++xFOAAAAAHOK7CVY a9eulSS1bdvWo9/p06dVqtSF7FS6dGmdOXPGaG0AAAAAQlNkA8j06dM1dOhQr/0uu+wyZ+g4c+aM ypQpY7I0AAAAACEqsgFkw4YN6tu3r/MSqfyXSsXHxzsvuzp06JAqVqwYhQoBAAAABKvIBhCbzeb8 5+h2SE5OVkZGhnJzc5WZman69etHp0gAAAAAQSmyAcSfPn36aMWKFWrZsqVWrFihBx54INolAQAA AAhAkX4KloP7U6xatmyp1atXR6cYAAAAACG7KM+AAAAAALg4EUAAAAAAGEMAAQAAAGAMAQQAAACA MQQQAAAAAMYQQAAAAAAYQwABAAAAYAwBBAAAAIAxBBAAAAAAxhBAAAAAABhDAAEAAABgDAEEAAAA gDEEEAAAAADGEEAAAAAAGEMAAQAAAGAMAQQAAACAMQQQAAAAAMYQQAAAAAAYQwABAAAAYAwBBAAA AIAxBBAAAAAAxhBAAAAAABhDAAEAAABgDAEEAAAAgDEEEAAAAADGEEAAAAAAGEMAAQAAAGAMAQQA AACAMaWiXQBwqbNYLCGNZ7fbw1wJAABA5BFAgCLgo00Hghq+U5NqEaoEAAAgsrgECwAAAIAxBBAA AAAAxhBAAAAAABhDAAEAAABgDAEEAAAAgDEEEAAAAADGEEAAAAAAGMN7QOAVL8cDAABAJBBA4BMv xwMAAEC4cQkWAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAA9d8NYQAAE6FJREFUYwgg AAAAAIwhgAAAAAAwhgACAAAAwBgCCAAAAABjCCAAAAAAjCGAAAAAADCmVLQLuJRZLJaQxrPb7WGu BN6wfgAAAMKPABJlH206ENTwnZpUi1Al8Ib1AwAAEF5cggUAAADAGAIIAAAAAGMIIAAAAACM4R6Q i1AoN0ebujG6KNcGAACA6COAXISK8o3RRbk2AAAARB+XYAEAAAAwhgACAAAAwBgCCAAAAABjCCAA AAAAjCGAAAAAADCGAAIAAADAGAIIAAAAAGMIIAAAAACMIYAAAAAAMIYAAgAAAMAYAggAAAAAY0pF uwCYYbFYol0CAAAAQAC5VHy06UBQw3dqUi1ClQAAAOBSxiVYAAAAAIwhgAAAAAAwhgACAAAAwBgC CAAAAABjCCAAAAAAjCGAAAAAADCGAAIAAADAGN4DEka87A8AAADwjwASRrzsDwAAAPCvyF6CtWbN GnXs+P/au9/Yqu76D+CfBsofsyxrBWRmxvtATbMS1GUTHXRop0EYk2mcLkidkcREgrhEywjNUONc lhDFgFtkCRgJ+G8PnD7ZjC4gaUahjQoKwemWTgNu1bVGOxfSUX4Plt5f6XoP/fM957bl9UpI1p7e 8z6c887hfHbP6f1INDY2xurVq+PMmTPlZSdPnozm5uZYsmRJrFmzJjo6Oqq4pQAAwFhN2QGkra0t 7rvvvujq6orbb7897r///vKyvXv3xrp166Krqyuamppi//79VdxSAABgrKbsLVjt7e3l/960aVP8 4Ac/KH999uzZ+MY3vhHz5s2L9evXx8aNGzPXVSqV8tpMAABgHKbsADLcyy+/HDfccEP5697e3qir q4uIiMWLF0dPT0/m67u7uysuM5wAAEBxpuwtWMM9/vjjsXnz5vLXFy5ciNmzX5+d5syZEwMDA9Xa NAAAYBym/ABy+vTp6OvrizvvvLP8vdra2vLQMTAwEHPnzq3W5gEAAOMwpQeQc+fOxb59+6Ktre2y 7y9YsKB821VPT0/U19dXY/MAAIBxmrIDSHt7e2zfvj0efPDBmDdv3mXLGhsb4+DBg9Hf3x+HDh2K hoaGKm0l1VBTUzOhPwAAVN+UfQh969atcf78+bjxxhvL3xt6mHzDhg2xbdu2OHDgQCxatCh27dpV pa2kWnzoIwDA9DRlB5Bnnnmm4rJly5bF4cOHC9waAAAghSl7CxYAADDzTNl3QLh6eD6DqW6iHb10 6VLiLQGA6c8AQtV5noPpQE8BIA23YAEAAIUxgAAAAIUxgAAAAIXxDAhXDQ+7j5+HrwGA1AwgXDU8 RDwx9hsAkJJbsAAAgMIYQAAAgMIYQAAAgMIYQAAAgMIYQAAAgMIYQAAAgMIYQAAAgML4HBBIzAce AgBUZgCBxHxwHwBAZW7BAgAACmMAAQAACmMAAQAACmMAAQAACmMAAQAACmMAAQAACmMAAQAACmMA AQAACmMAAQAACmMAAQAACmMAAQAACjO72hsATExNTc24X3Pp0qUctgQAYOwMIDBN/er3/xjXz696 7/U5bQkAwNi5BQsAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiM AQQAACiMAQQAACiMAQQAACjM7GpvAFCcmpqaam8CAHCVM4DAVeRXv//HuH5+1Xuvz2lLAICrlVuw AACAwhhAAACAwhhAAACAwngGBEhuIg+7X7p0KYctqa6ZtB8m+gsMpurfB4DqMYAAyXnY/XUzbT/M tL8PANXhFiwAAKAwBhAAAKAwbsECpqWZ+kzCTHpuBABGYwABpq2Z+EzCTPw7AcBwbsECAAAKYwAB AAAKYwABAAAK4xkQABiDmfqLDwCKZgABgDHySwIAJs8tWAAAQGEMIAAAQGHcggUwzfnwwqnN8QG4 nAEEYJrzXMLU5vgAXM4tWAAAQGEMIAAAQGEMIAAAQGE8AwIAAOPkw0knzgACAAAT4JdMTIxbsAAA gMIYQAAAgMK4BQvgKjTRe5eLyHF/9PgVdS+6e97tgwj7gMkzgABchYq6b9n90cVxTItjH9gHTI5b sAAAgMIYQAAAgMIYQAAAgMJ4BgSYEqbyQ9FQND0FZjIDCDAlTNUHaCeTBRPlAV9gJnMLFgAAUBgD CAAAUBi3YAEwrXleYuKm6r7zQXdT/8MlJ8IHk76uiP1Q5HGdyDEygAAw7U3kmQnPWUztZ02m8rYV Zao+GzfVc6a6qby/i9q2aXkL1smTJ6O5uTmWLFkSa9asiY6OjmpvEgAAMAbTcgDZu3dvrFu3Lrq6 uqKpqSn2799f7U0CAADGoKa7u3va3VzX3NwcP/3pT2PhwoXxwgsvxMaNG+M3v/lNxZ8vlUrFbRwA AFzFuru7M5dPy2dAent7o66uLiIiFi9eHD09PZk/f6WdMFalUmnc65rKrykyy2u8xmu8xmumx2uK zPIar/GamfmaK5mWt2BduHAhZs9+fXaaM2dODAwMVHmLAACAsZiWA0htbW156BgYGIi5c+dWeYsA AICxmJYDyIIFC8q3XfX09ER9fX2VtwgAABiLaTmANDY2xsGDB6O/vz8OHToUDQ0N1d6kilLfM1ft nCKz5MiRI0fO9MgpMkuOHDlTP+dKpuUAsmHDhnjqqadi2bJl8dRTT8UXvvCFam8SAAAwBtPyt2At W7YsDh8+XO3NAAAAxmlafg4IAAAwPU3LW7AAAIDpyQACAAAUxgACAAAUxgACAAAUxgACAAAUxgAC AAAUxgACAAAUxgAyBidPnozm5uZYsmRJrFmzJjo6OnLJ+e1vfxsf+chHorGxMVavXh1nzpzJJWfI 0aNHo1Qq5bb+ixcvxtatW+Pd7353rFixIo4ePZpLztmzZ2Pt2rXR2NgY69ati+eeey7Zunt7e+PW W299w/dTd6JSTupOVMoZkqoTlXJSd6JSTupOZB2HlF3IyknZhbGsK0UXsnJSdiErJ2UXzpw5E3fc cUcsWbIk7rrrrvjLX/5SXpayB1k5KXuQlTMkRQ+yclL2ICsnj38nRts3eVwvjJaTx/VC1rFOeb0w 2rryuF4YLSd1D86fPx+lUumyP0NSdiErJ2UXsnKGpOqCAWQM9u7dG+vWrYuurq5oamqK/fv355LT 1tYW9913X3R1dcXtt98e999/fy45Q/bs2ZPr+vft2xfXXnttdHR0xLZt22LHjh255LS2tkZra2uc OnUqvvSlL0Vra2uS9f7973+PDRs2xPnz59+wLGUnsnJSdiIrZ0iKTmTlpOxEVk7qTmQdh5RdyMpJ 2YWxrCtFF7JyUnYhKydlF77yla9ES0tLdHZ2xkc/+tH46le/Wl6WsgdZOSl7kJUzJEUPsnJS9iAr J49/J0bbN3lcL4yWk8f1QtaxTnm9MNq68rheGC0ndQ/+8Ic/xMqVK6O7u7v8Z0jKLmTlpOxCVs6Q VF0wgIzB2bNnY/369TFv3rxYv359PP/887nktLe3xx133BHz58+PTZs25ZYT8foEW1NTk9v6IyKe fPLJWLt2bcyfPz/Wrl0bR44cySXn+eefj+XLl8esWbNi5cqV8eyzzyZZ79133x2f+tSnRl2WshNZ OSk7kZUTka4TWTkpO5GVk7oTWcchZReyclJ24UrrStWFrJyUXcjKSdmFJ598Mu65556YP39+3Hvv vbn1ICsnZQ+yciLS9SArJ2UPsnJSnxMq7ZvU1wuVclJfL2Qd65TXC5XWlfp6oVJO6h788Y9/jMbG xlGXpexCVk7KLmTlRKTtggFkDHp7e6Ouri4iIhYvXhw9PT25Z7788stxww035Lb+PXv2xJYtW3Jb f0TE3/72tzh8+HDcdNNNsWrVqqS3Rg3X0NAQzzzzTAwODsbx48fjXe96V5L1/vznP4/Pfe5zoy5L 2YmsnOEm24kr5aTqRFZOyk5k5eTViYg3Hoe8zg9Zxzvl+WG0deVxfhiZk9f5YWROHl343//+F7t3 746bb765/L08ejBaznCpelApJ3UPRsvJowej5aTuQaV9k7oHYzkGKXqQlZOyB5XWlboHlXJS9+D0 6dNx9OjRWLp0aaxdu/aydwxSdiErZ7jJduFKOSm7MDvJWma4CxcuxOzZr++qOXPmxMDAQO6Zjz/+ eGzevDmXdQ/dW7lixYpc1j/k1Vdfjf7+/jh27Fg8/fTT8cADD8SPfvSj5Dnbt2+Pe++9N/r7++Oa a66JH/7wh0nWe/3111dclrITWTnDTbYTWTkpO5GVk7ITWTl5dSLijcchr/ND1vFOeX4Yua68zg8j c/I6P4zMSd2Fvr6+WLVqVbzyyivxrW99q/z91D2olDNcih5Uykndg0o5qXtQKSdlD7L2TcoejPUY TLYHWTkpe5C1rpQ9yMpJfT549tln46GHHoqVK1dGV1dXtLW1xaFDhyIibReycoabbBeyclKfE7wD Mga1tbXl4gwMDMTcuXNzzTt9+nT09fXFnXfemcv6d+/enfu7HxERs2bNii1btsTcuXPjwx/+cPzp T3/KJedrX/taPPbYY/Hcc8/F7t2745vf/GYuOcPpxMRM906Mdhzy6ELW8U7ZhdHWlUcXRsvJowuj 5aTuQl1dXZw4cSK+/vWvx86dO8vfT92DSjlDUvWgUk7qHlTKSd2DSjkpe5C1b1L2YCzHIEUPsnJS 9iBrXSl7kJWT+nzQ0dERzc3NMWvWrLjlllvi1KlT5WUpu5CVMyRFF7JyUp8TDCBjsGDBgvJbZz09 PVFfX59b1rlz52Lfvn3R1taWW0ZXV1e0tLSUf4tBXr8Ja+HChXHhwoWIiBgcHCz/n4DUXnjhhfjA Bz4Qs2bNiqampvjrX/+aS85wOjEx07kTlY5D6i5kHe+UXai0rtRdqJSTuguVcvI6P3zsYx+Lf//7 3+Wv8zonjMyJyOecMDInr3PCyJy8zgkjc1L2IGvfpOzBlY5Bqh5k5aTsQda6UvYgKyfP64XBwcGo ra0tf53XOWFkTkQ+54SROanPCQaQMWhsbIyDBw9Gf39/HDp0KBoaGnLJaW9vj+3bt8eDDz4Y8+bN yyUjIt7w2w0q3Us4Wbfddlv8+Mc/josXL8avf/3reN/73pdLzjvf+c7o7OyM1157LU6cOBFvf/vb c8kZTicmZrp2Ius4pOxCVk7KLmStK2UXsnJSdiErJ2UXVq5cGcePH4/BwcE4cuRI3HjjjeVlKXuQ lZOyB1k5KXuQlZOyB1k5KXuQtW9S9iArJ2UPsnJS9iBrXSl7kJWT+t+G2267LX73u9/F4OBgdHR0 xPLly8vLUnYhKydlF7JyUl8n1HR3d1+a1BquAsePH49t27ZFT09PLFq0KHbt2hXvec97kufceuut b/iVonldCA4plUq5ZfT398eWLVvixIkT8Y53vCMeffTReOtb35o858yZM9Ha2hrd3d1RKpVi586d l/3DM1mj7aM8OjFaTh6duNIxT9WJ0daTRydGy0ndiazjkLILWTkpuzDWdU22C1k5KbuQlZOyC52d ndHW1hbnzp2LhoaG+M53vlO+gEnZg6yclD3Iyhlusj3IyknZg6ycvP6dGLlv8rpeGJmT1/VC1rFO eb0wcl15XS+MzEndg2PHjsUDDzwQL774YixdujS+973vld/pSNmFrJyUXcjKGS5FFwwgAABAYdyC BQAAFMYAAgAAFMYAAgAAFMYAAgAAFMYAAgAAFMYAAgAAFMYAAgAAFMYAAgAAFMYAAgAAFMYAAgAA FMYAAgAAFMYAAgAAFGZ2tTcAgJnvn//8Zzz88MPR3t4e/f39UV9fHx/60Idi69atcc0111R78wAo kHdAAMjVf//73/j0pz8dnZ2dsX///jhx4kTcc889ceDAgfjiF79YyDYcPXo0PvnJTxaSBUC2mu7u 7kvV3ggAZq6HHnooHnvssdi5c2fcfffdVdmGUqkUERHd3d1VyQfg/3kHBIBcHTlyJCIiVqxYUfFn XnvttdixY0fcfPPNcdNNN8WOHTvi4sWL5eWlUqk8RGR9/fTTT8fy5ctj2bJl8cQTT5SXDf85AKrL AAJArl588cWIiFi0aFHFn/nud78bBw4ciE2bNsXmzZvjwIEDsWvXrnFnnTp1Kh5++OF46aWX4tvf /nZEXP6uh3dAAKrPAAJArgYHByMi4tKlynf8/vKXv4yIiI9//ONx1113RUTEL37xi3FntbS0xPvf //6IiPjXv/417tcDkD8DCAC5Gnrn46WXXqr4M729vRERcd1118V111132fdGyhpkFixYELW1tRER 8eqrr05oewHIlwEEgFw1NTVFRMSxY8cq/kxdXV1ERPT19UVfX19ERLz5zW8uL3/Tm94UEa8/KzK0 HIDpyQACQK6+/OUvx9ve9rbYs2dP/PnPf45XXnklHnnkkSiVStHS0hIREatXr46IiCeeeKJ869Un PvGJ8joWLlwYERGdnZ2xZ8+ecW/D0DDzn//8Z1J/FwAmzwACQK7q6+vjZz/7WSxdujQ+85nPxC23 3BI/+clPoqWlJb7//e9HRERra2t89rOfjUcffTQeeeSR+PznPx9btmwpr6O1tTXe8pa3xPbt2+OD H/zguLehtbU1rr322ti4cWOqvxYAE+RzQAAAgMJ4BwQAACiMAQQAACiMAQQAACiMAQQAACiMAQQA ACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACiMAQQAACjM /wGuYHgq+Y9UZwAAAABJRU5ErkJggg== ------=_NextPart_000_0000_01CB6F6B.332B4A50--