Uses the file cardio_train.csv downloaded from https://www.kaggle.com/sulianova/cardiovascular-disease-dataset
Implements a function analyse(gender, age) where gender is one of ‘F (for female) or ‘M’ (for male), and where age is any integer for which we have data in the file (nothing needs to be done if that is not the case).
We assume that all years have 365 days; in particular, someone who is 365 days old is 0 year old, and someone who is 366 days old is 1 year old.
We ignore records for which at least one of these conditions holds:
height < 150 or height > 200
weight < 50 or weight > 150
ap_hi < 80 or ap_hi > 200
ap_lo < 70 or ap_lo > 140
For each of both classes “cardio problem” and “no cardio problem” (as given by the cardio attribute), we create 5 bins/categories for height, weight, ap_hi, ap_lo, of equal width, that span between smallest value and largest value for the attribute in the category. For instance, suppose that gender is ‘F’ and age is 48.
Suppose that for the category “cardio problem”, the shortest woman aged 48 is 150cm tall, and the tallest woman aged 48 is 200cm tall. Then each of the 5 categories for the class “cardio problem” and for the attribute “height” spans 10cm.
Suppose that for the class “no cardio problem”, the shortest woman aged 48 is 158cm tall, and the tallest woman aged 48 is 193cm tall. Then each of the 5 categories for the class “no cardio problem” and for the attribute “height” spans 7cm.
To avoid bounday issues, add 0.1 to the maximum value (so with the previous example, the maximum heights would be considered to be 200.1 and 193.1, respectively). This applies to each of the 4 attributes height, weight, ap_hi and ap_lo.
For each attribute and for each of its possible values, we compute the ratio of
the frequency of people under consideration with a “cardio problem” having that value for that attribute, with
the frequency of people under consideration with “no cardio problem” having that value for that attribute.
Continuing the previous example:
Suppose that there are 100 woman aged 48 who have a “cardio problem” and 20 of those are at most 160cm tall.
Suppose that there are 150 woman aged 48 who have “no cardio problem” and 50 of those are at most 165cm tall.
Then the ratio for the value “category 1” of the attribute “height” is 0.2 / 0.3…3…
We keep only ratios that are strictly greater than 1 and order them from largest to smallest. A ratio might be infinite (see second sample test). In case two ratios are exactly the same, their order is determined by the order of the corresponding attributes in the csv file (first is height, last is being active or not), and in case the attributes are the same, their order is determined by the rank of the category (first is 1, last is 5; for booleans, False comes before True).
We format ratios with 2 digits after the decimal point. After a ratio, the output is one of:
Height in category [1-5] (1 lowest, 5 highest)
Weight in category [1-5] (1 lowest, 5 highest)
Systolic blood pressure in category [1-5] (1 lowest, 5 highest)
Diastolic blood pressure in category [1-5] (1 lowest, 5 highest)
Cholesterol in category [1-3] (1 lowest, 3 highest)
Glucose in category [1-3] (1 lowest, 3 highest)
Smoking/Not smoking
Drinking/Not drinking
Not being active/Being active
You are NOT allowed to use pandas. If you do, then your submission will NOT be assessed and you will score 0 to the quiz.