i will attached a PDF file with the instruction, please use SAS to do the project

Introduction:
We have selected BoatDemand data for analysis. First, we have identified response and explanatory variable. We have run summary statistics (mean, standard deviation etc.). we have selected explanatory variable based on domain and amount of association with response variable. We have transformed data by logarithmic but didn’t get ant significant impact transformation on predictability of predictors.
We have built multiple linear regression using number of trips as response variable. Our regression model is statistically significant to predict number of trips. Now we can have information regarding drivers of number of trips. That will ultimately help us to make informed decision. We can increase number of trips by increasing positively associated predictors.
Summary Statistics:
    Variable
    Mean
    Std Dev
    Minimum
    Maximum
    N
    costC
income
costS
costH
trips
    55.423
3.852
59.928
55.990
2.244
    46.682
1.851
46.376
46.133
6.292
    4.340
1.000
4.767
5.700
0
    493.770
9.000
491.547
491.049
88.000
    659
659
659
659
659
Co
elation coefficient:
Co
elation coefficients are utilized to evaluate the strength of a connection between two persistent factors. Pearson, Spearman, and Kendall are the three strategies for computing co
elation coefficients. By and large, Pearson co
elation is utilized in direct relapse. The coefficients of co
elation range from - 1 to 1. A negative co
elation coefficient (esteem between - 1 and 0) shows the backwards relationship between factors. Positive co
elation coefficients (coefficient values going from 0 to 1) suggest an immediate relationship between factors.
All variables are having negative co
elation coeffcient except quality. Co
elation coefficient between trips and costC are negative and very low. Co
elation coefficient between trips and costS are negative and low. Co
elation coefficient between trips and costH are negative and very low. Co
elation coefficient between trips and quality are positive and moderate.
    Pearson Co
elation Coefficients, N = 659
     
    trips
    costC
    -0.04221
    costS
    -0.12370
    costH
    -0.02051
    income
    -0.06003
    quality
    0.3863
Regression Model:
Simple linear regression
A factual technique called simple linear regression permits us to characterize and inspect relationships between two factors. One indicator and one reaction variable are utilized in simple linear regression. A reaction variable might be addressed by the main variable (quantitative ward variable). An illustrative variable or an autonomous variable might be utilized to depict the subsequent variable. Indicators are one more word for informative factors.
The dispersion of two factors by every perception is alluded to as a dissipate plot. Drawing a dissipate plot permits us to see the association between two factors rapidly. In the dissipate plot, each dab addresses a perception. We might fit a trendline in a dissipate plot, which shows the observational association among reliant and free factors. The contrast among genuine and projected beta coefficients is addressed by the standard mistake of assessed measures in the fundamental linear regression model.
σest = √σ(y - ŷ)2/n
where:
y: The actual value
ŷ: The anticipated value
n: The all-out number of records
On account of model precipitation and collect yield, yield relies upon the proportion of precipitation; yield is the reliant variable, and precipitation is the informative variable. Regression examination is separated into two kinds in view of the quantity of accessible ward factors, like linear regression and multiple linear regression.
Linear Regression Significance Test:
The coefficient of the regression model in condition 1 is the focal point of the basic preliminary of direct regression. Where is the tendency and steady If it isn't equivalent to nothing, there's a critical connection between the free and ward components hypothesis for assessing direct regression importance as Ho: = 0, Ha: 0. The invalid hypothesis expresses that the coefficient approaches zero, however the elective hypothesis expresses that the coefficient isn't zero. It is a perspective wherein the worth of condition 1 is zero; by then, condition 1 tends to a steady inspiration for each independent worth. Mistake: Residual (e) is an aftereffect of an examination between the real worth of the dependent(target) variable (y) and the normal worth () utilizing a model.
Null Hypothesis: Beta Coefficient isn't not different from zero.
Elective hypothesis: : Beta Coefficient is different from zero.
We have conducted multiple linear regression analysis to predict number of trips. What are the factors behind increase or decrease in number of trips? We have set significance level at 5 percent. We have portioned data using 80:20 split.
Dependent variable:
Independent elements impact the adjustment of the dependent variable. Assume we check the age and tallness of youths out. The dependent variable will be tallness, and the independent variable will be age. Whenever we as a whole know, as a youth progresses in years, their size will in general ascent. The age of a kid is independent of any remaining variables.
Independent variable:
Assume we check the race and tallness of kids out. The...