Intro to Python, Class 2 Solutions

Challenge-import

Example 1:

urllib.request.urlretrieve("https://raw.githubusercontent.com/fredhutchio/R_intro/master/extra/clinical.tsv", "data/clinical.tsv")
example1 = pd.read_csv("../data/clinical.tsv", sep="\t")

Example2:

urllib.request.urlretrieve("https://raw.githubusercontent.com/fredhutchio/R_intro/master/extra/clinical.txt", "data/clinical.txt")
example2 = pd.read_csv("../data/clinical.txt", sep=" ")

Challenge-typo

If you have a typo in a column name, like clinical_df.tumorstage, the result is KeyError: tumorstage, found at the bottom of the TraceBack. This is cryptic until you understand that the “Key” is the column name!

Challenge-order

If you reverse the order of the columns, clinical_df[["vital_status", "tumor_stage"]], they return results in a different order. This means that columns don’t have to be entered in same order as original data, but the order of columns listed will affect order in the output.

Challenge-last

There are multiple solutions.

clinical_df[-10:]

Challenge-location

The numbers are being interpreted as labels, and there are no labels with those names.

One possible solution:

clinical_df.iloc[2, 6]

However, it’s not not clear what the intention was behind the code, so it’s not really possible to correct it!

Challenge-100

Two options:

clinical_df.loc[6732:, ["vital_status", "days_to_death"]]
clinical_df.iloc[-100:, [3,5]]

Challenge-object

Try this on any of the columns of type “object”:

clinical_df.site_of_resection_or_biopsy.describe()

In the output, the object (categorical) data are described with:

• count: total observations
• unique: number of categories
• top: most frequently occurring category
• freq: number of observations in the most frequently occurring category

Challenge-deviation

clinical_df.days_to_death.std()

Challenge-ref

This challenge is a bit easier to assess if you print only a portion of the objects using head:

• clinical_df.head() has been modified because ref_clinical_df referenced it
• ref_clinical_df.head() was directly altered in our code
• true_copy_clinical_df.head() is the copy of original, and remains unaltered

Challenge-first-five

Using iloc:

clinical_df.iloc[:5 , [0, 1, 2, 3, 13, 19]]

Using loc:

clinical_df.loc[:4 , ["primary_diagnosis", "tumor_stage", "age_at_diagnosis", "vital_status", "gender", "disease"]]

Challenge-summary

There are multiple ways of addressing this problem. One possible solution:

# group by disease
grouped_data = clinical_df.groupby("disease")
# group by vital status and summarize
grouped_data.vital_status.describe()

Challenge-last

test_data[2]
test_data[-1]

Challenge-compare

# compare minimum age at diagnosis between original clinical and true copy of clinical
diff_clinical_true = clinical_df.age_at_diagnosis.min() - true_copy_clinical_df.age_at_diagnosis.min()
# compare minimum age at diagnosis between original clinical and referenced clinical
diff_clinical_ref = clinical_df.age_at_diagnosis.min() - ref_clinical_df.age_at_diagnosis.min()
# print difference between original and true copy
print(diff_clinical_true)
# print difference between original and referenced data
print(diff_clinical_ref)