ISE529 Module 1 Homework Solution

ISE-529 Predictive Analytics
1. For this problem, we will be using the file "Cars Data.csv"
a. Load the file Cars Data.csv into a dataframe named cars and display the first 10 rows of the dataframe.
In [117… import pandas;

cars = pandas.read_csv(filepath_or_buffer = "Cars Data.csv", header = 3); print(cars.head(10));
Make Model DriveTrain Origin Type Cylinders
0 Acura 3.5 RL 4dr Front Asia Sedan 6.0 1 Acura 3.5 RL w/Navigation 4dr Front Asia Sedan 6.0
2 Acura MDX All Asia SUV 6.0
3 Acura NSX coupe 2dr manual S Rear Asia Sports 6.0
4 Acura RSX Type S 2dr Front Asia Sedan 4.0 5 Acura TL 4dr Front Asia Sedan 6.0
6 Acura TSX 4dr Front Asia Sedan 4.0
7 Audi A4 1.8T 4dr Front Europe Sedan 4.0
8 Audi A4 3.0 4dr Front Europe Sedan 6.0
9 Audi A4 3.0 convertible 2dr Front Europe Sedan 6.0
Engine Size (L) Horsepower Invoice Length (IN) MPG (City)
0 3.5 225 $39,014 197 18 1 3.5 225 $41,100 197 18
2 3.5 265 $33,337 189 17
3 3.2 290 $79,978 174 17
4 2.0 200 $21,761 172 24 5 3.2 270 $30,299 186 20
6 2.4 200 $24,647 183 22
7 1.8 170 $23,508 179 22
8 3.0 220 $28,846 179 20
9 3.0 220 $38,325 180 20
MPG (Highway) MSRP Weight (LBS) Wheelbase (IN)
0 24 $43,755 3880 115 1 24 $46,100 3893 115
2 23 $36,945 4451 106
3 24 $89,765 3153 100
4 31 $23,820 2778 101 5 28 $33,195 3575 108
6 29 $26,990 3230 105
7 31 $25,940 3252 104
8 28 $31,840 3462 104 9 27 $42,490 3814 105
b. Use the describe() function to produce a numerical summary of the variables in the dataset. (10 points)
In [118… print(cars.describe());
Cylinders Engine Size (L) Horsepower Length (IN) MPG (City) count 426.000000 428.000000 428.000000 428.000000 428.000000 mean 5.807512 3.196729 215.885514 186.362150 20.060748 std 1.558443 1.108595 71.836032 14.357991 5.238218 min 3.000000 1.300000 73.000000 143.000000 10.000000 25% 4.000000 2.375000 165.000000 178.000000 17.000000
50% 6.000000 3.000000 210.000000 187.000000 19.000000 75% 6.000000 3.900000 255.000000 194.000000 21.250000 max 12.000000 8.300000 500.000000 238.000000 60.000000
MPG (Highway) Weight (LBS) Wheelbase (IN) count 428.000000 428.000000 428.000000 mean 26.843458 3577.953271 108.154206 std 5.741201 758.983215 8.311813 min 12.000000 1850.000000 89.000000 25% 24.000000 3104.000000 103.000000 50% 26.000000 3474.500000 107.000000 75% 29.000000 3977.750000 112.000000 max 66.000000 7190.000000 144.000000
c. Use the groupby() and size() functions to create a table of the number of observations in the dataset of each car Make (e.g., Acura, Audi, etc.) (15 points)
In [119… print(cars.groupby("Make").size());
Make
Acura 7 Audi 19
BMW 20
Buick 9
Cadillac 8 Chevrolet 27
Chrysler 15
Dodge 13
Ford 23 GMC 8
Honda 17
Hummer 1
Hyundai 12 Infiniti 8
Isuzu 2
Jaguar 12
Jeep 3 Kia 11
Land Rover 3
Lexus 11
Lincoln 9 MINI 2
Mazda 11
Mercedes-Benz 26
Mercury 9 Mitsubishi 13
Nissan 17
Oldsmobile 3
Pontiac 11 Porsche 7
Saab 7
Saturn 8
Scion 2 Subaru 11
Suzuki 8
Toyota 28
Volkswagen 15 Volvo 12 dtype: int64
d. Use the corr() function to create a correlation matrix of the numeric attributes in the cars dataset. Use the "pearson" correlation coefficient algorithm (15 points)
In [120… print(cars.corr(method = "pearson"));
Cylinders Engine Size (L) Horsepower Length (IN)
Cylinders 1.000000 0.908002 0.810341 0.547783
Engine Size (L) 0.908002 1.000000 0.787435 0.637448 Horsepower 0.810341 0.787435 1.000000 0.381554
Length (IN) 0.547783 0.637448 0.381554 1.000000
MPG (City) -0.684402 -0.709471 -0.676699 -0.501526
MPG (Highway) -0.676100 -0.717302 -0.647195 -0.466092 Weight (LBS) 0.742209 0.807867 0.630796 0.690021
Wheelbase (IN) 0.546730 0.636517 0.387398 0.889195
MPG (City) MPG (Highway) Weight (LBS) Wheelbase (IN) Cylinders -0.684402 -0.676100 0.742209 0.546730
Engine Size (L) -0.709471 -0.717302 0.807867 0.636517
Horsepower -0.676699 -0.647195 0.630796 0.387398
Length (IN) -0.501526 -0.466092 0.690021 0.889195 MPG (City) 1.000000 0.941021 -0.737966 -0.507284
MPG (Highway) 0.941021 1.000000 -0.790989 -0.524661
Weight (LBS) -0.737966 -0.790989 1.000000 0.760703
Wheelbase (IN) -0.507284 -0.524661 0.760703 1.000000
e. Add a new attribute to the dataframe called HP_Groups which has one of the following values:
"Low": Horsepower is <= 200
"Medium": Horsepower is > 200 and <= 300
"High": Horsepower is > 300
Hint: create a function that takes a number and return one of the three bins and then use the apply method
hpList = []; for index, row in cars.iterrows():
temp = ""; if row["Horsepower"] <= 200:
temp = "Low"; elif row["Horsepower"] > 200 and row["Horsepower"] <= 300:
temp = "Medium"; else:
temp = "High"; hpList.append(temp);

cars["HP_Groups"] = hpList; print(cars);
In [121…
Make Model DriveTrain Origin Type Cylinders
0 Acura 3.5 RL 4dr Front Asia Sedan 6.0 1 Acura 3.5 RL w/Navigation 4dr Front Asia Sedan 6.0
2 Acura MDX All Asia SUV 6.0
3 Acura NSX coupe 2dr manual S Rear Asia Sports 6.0
4 Acura RSX Type S 2dr Front Asia Sedan 4.0 .. ... ... ... ... ... ...
423 Volvo S80 2.9 4dr Front Europe Sedan 6.0
424 Volvo S80 T6 4dr Front Europe Sedan 6.0
425 Volvo V40 Front Europe Wagon 4.0 426 Volvo XC70 All Europe Wagon 5.0
427 Volvo XC90 T6 All Europe SUV 6.0
Engine Size (L) Horsepower Invoice Length (IN) MPG (City) 0 3.5 225 $39,014 197 18
1 3.5 225 $41,100 197 18
2 3.5 265 $33,337 189 17
3 3.2 290 $79,978 174 17 4 2.0 200 $21,761 172 24 .. ... ... ... ... ...
423 2.9 208 $35,542 190 20
424 2.9 268 $42,573 190 19 425 1.9 170 $24,641 180 22
426 2.5 208 $33,112 186 20
427 2.9 268 $38,851 189 15
MPG (Highway) MSRP Weight (LBS) Wheelbase (IN) HP_Groups
0 24 $43,755 3880 115 Medium
1 24 $46,100 3893 115 Medium
2 23 $36,945 4451 106 Medium 3 24 $89,765 3153 100 Medium
4 31 $23,820 2778 101 Low .. ... ... ... ... ...
423 28 $37,730 3576 110 Medium 424 26 $45,210 3653 110 Medium
425 29 $26,135 2822 101 Low
426 27 $35,145 3823 109 Medium
427 20 $41,250 4638 113 Medium
[428 rows x 16 columns]
f. Using a Seaborn, create side-by-side boxplots showing the MPG (City) for each of the three HP_Groups
In [122… import seaborn;

seaborn.boxplot(x = cars["HP_Groups"], y = cars["MPG (City)"]);

2. Function Definition
A common metric of health is the Body Mass Index (BMI), which is calculated simply as the weight in kilograms divided by the height in meters squared (BMI = kg/m2)
a. Create a function called calculate_bmi() that takes height in inches and weight in pounds as parameters and returns the BMI. Use theconversions inches ∗ 0.025 = m and pounds ∗ 0.453592 = kg. Test the function with a height of 72" and a weight of 190 lbs. (15 points)
def calculate_bmi(inches, pounds):
kg = pounds * 0.453592; m = inches * 0.025; return kg / pow(m, 2);

print(calculate_bmi(72, 190));
In [123…
26.59953086419753
b. General standard categories for BMI are given by:
BMI < 18.5: Underweight
BMI 18.5 - 24.9: Normal weight
BMI 25 - 29.9: Overweight
BMI 30 or greater: Obese
Create a function called determine_weight_category that takes height in inches and weight in pounds as parameters and returns the category the person falls into. Test your function with the following values: (15 points)
Height 60" / Weight 160 lbs
Height 68" / Weight 160 lbs
Height 72" / Weight 160 lbs
def determine_weight_category(inches, pounds):
bmi = calculate_bmi(inches, pounds); print(bmi); if bmi < 18.5:
return "Underweight"; elif bmi >= 18.5 and bmi < 25: return "Normal weight"; elif bmi >= 25 and bmi < 30:
return "Overweight"; else:
return "Obese";

print(determine_weight_category(60, 160)); print(determine_weight_category(68, 160)); print(determine_weight_category(72, 160));
In [124…
32.25543111111111
Obese
25.112359861591692 Overweight
22.399604938271604
Normal weight
c. You want to create a plot of the BMI for a 180-pound individual at various heights from 60" to 84" (in one-inch increments). Create a vector ‘heights’ to hold the various heights from 60 to 84 and write a for-loop to call your calculate_bmi() function for each value of heights. Then, use these two vectors to create a scatter plot showing the relationship. Give the chart a title of ‘BMI at Various Heights for a 180-Pound Individual’.
For full credit, label the axes and give the chart a title.
heights = []; bmiList = [];
for a in range(60, 85): heights.append(a); for h in heights:
bmiList.append(calculate_bmi(h, 180));

df = pandas.DataFrame(
{
"height" : heights,
"bmi" : bmiList
}
)
seaborn.scatterplot(data = df, x = "height", y = "bmi").set(title = "BMI at Various Heights for a 180-Pound Individual");
In [125…

d. Repeat part c from above but without using a loop
def calculate_bmi_2(inches):
kg = 180 * 0.453592; m = inches * 0.025; return kg / pow(m, 2);

series_heights = pandas.Series(heights);
series_bmi = series_heights.apply(calculate_bmi_2);

df2 = pandas.DataFrame(
{
"height" : series_heights,
"bmi" : series_bmi
}
)
seaborn.scatterplot(data = df2, x = "height", y = "bmi").set(title = "BMI at Various Heights for a 180-Pound Individual");
In [126…

3) Demographics Dataset Analysis
a. For this problem, load the file "demographics.csv" into a DataFrame called "demos". Uset the "type" function to verify that it is a dataframe
In [127… demos = pandas.read_csv(filepath_or_buffer = "demographics.csv"); print(type(demos)); print(demos);
<class 'pandas.core.frame.DataFrame'>
CONT NAME Region Population Urban Pop Percentage
0 91 BAHAMAS AMR 323063 0.90 1 91 BELIZE AMR 269736 0.49
2 91 CANADA AMR 32268243 0.81
3 91 COSTA RICA AMR 4327228 0.62
4 91 CUBA AMR 11269400 0.76 .. ... ... ... ... ...
192 96 NIUE WPR 1445 0.37
193 96 PAPUA NEW GUINEA WPR 5887138 0.13
194 96 TONGA WPR 102311 0.34 195 96 TUVALU WPR 10441 0.57
196 96 SAMOA WPR 184984 0.23
AdultLiteracyPct MaleSchoolPct FemaleSchoolPct GNI 0 NaN 0.85 0.88 16140.0
1 0.77 0.98 1.00 6510.0
2 NaN 1.00 1.00 30660.0
3 0.96 0.90 0.91 9530.0 4 1.00 0.96 0.95 NaN .. ... ... ... ...
192 NaN 0.99 0.98 NaN
193 0.57 0.79 0.69 2300.0 194 0.99 1.00 1.00 7220.0
195 NaN NaN NaN NaN
196 0.99 0.99 0.96 5670.0
PopPovertyPct
0 NaN
1 NaN
2 NaN 3 0.02
4 NaN .. ...
192 NaN 193 NaN
194 NaN
195 NaN
196 NaN
[197 rows x 10 columns]
b. For each continent, summarize the average (mean) city size
In [128… print(demos.groupby("CONT")[["Population"]].mean());
Population
CONT 91 3.182618e+07
92 1.984816e+07
93 1.695122e+07
94 1.707178e+07 95 8.953362e+07
96 4.030014e+06
c. Find the size of the largest country in each continent
In [129… print(demos.groupby("CONT")[["NAME", "Population"]].max());
NAME Population
CONT 91 UNITED STATES 298212895
92 VENEZUELA 186404913
93 UNITED KINGDOM 82689210
94 ZIMBABWE 131529669 95 YEMEN 1323344591 96 VANUATU 20155129
In [130… print(demos.groupby("CONT")[["MaleSchoolPct", "FemaleSchoolPct"]].mean());
MaleSchoolPct FemaleSchoolPct
CONT
91 0.932143 0.931429
92 0.930588 0.919412
93 0.942381 0.939286 94 0.734444 0.676222
95 0.885429 0.856000 96 0.933077 0.923077
In [131… print(demos.groupby("CONT")[["FemaleSchoolPct"]].mean().sort_values(["FemaleSchoolPct"], ascending = False));
FemaleSchoolPct
CONT
93 0.939286 91 0.931429
96 0.923077
92 0.919412
95 0.856000
94 0.676222