For this assignment, use the cc$\_$info.csv and transactions.csv files. Your

submission should include:

$1.$ The Python script.

$2.$ Sample output.

$3.$ An explanation of your approach and reasoning behind your code.

Each question is worth $10$ points, broken down as follows:

$$ Python script: $6$ points

$$ Output: $2$ points

$$ Code explanation and reasoning: $2$ points

Question $1$: Suspicious Transaction Analysis $(10$ points$)$

Company ABC suspects that transactions that exceed a certain threshold within a short period

of time might be fraudulent. Using the transactions DataFrame:

$1.$ Identify all transactions that are greater than $$300.$

$2.$ Sort these high$-$value transactions by the credit$\_$card column and then by date.

$3.$ For each credit$\_$card, calculate the time difference in days between consecutive

high$-$value transactions.

$4.$ Add a new column called suspicious to indicate whether a transaction is considered

suspicious. A transaction is considered suspicious if:

$$ It occurs within $3$ days of the previous high$-$value transaction for the same

credit$\_$card.

$5.$ Print the credit$\_$card, transaction$\_$dollar$\_$amount, date, suspicious

columns for all high$-$value transactions.

Hint: Use pd$.$to$\_$datetime$()$ to ensure date format compatibility and diff$()$ to calculate

consecutive differences in days.

Question $2$: Outlier Detection in Transaction Amounts $(10$ points$)$

Outlier detection can help identify unusual spending patterns. Using the transactions

DataFrame:

$1.$ Calculate the median and interquartile range $($IQR$)$ for the

transaction$\_$dollar$\_$amount column.

$2.$ Define transactions as "outliers" if their transaction$\_$dollar$\_$amount is either below

Q$1-1\mathrm{.}5*$ IQR or above Q$3+1\mathrm{.}5*$ IQR.

$3.$ Create a new column named is$\_$outlier in the transactions DataFrame. Set it to

True for outliers and False for non$-$outliers.

$4.$ Calculate and print the percentage of transactions flagged as outliers.

Hint: Use NumPy functions like np$.$percentile$()$ to calculate Q$1(25$th percentile$)$ and Q$3$

$(75$th percentile$).$

Question $3$: Daily Transaction Volume Analysis by Location $(10$ points$)$

Fraud patterns may also emerge from analyzing transaction volumes over time and by location.

Using both cc$\_$info and transactions DataFrames:

$1.$ Merge cc$\_$info and transactions on the credit$\_$card column.

$2.$ Group the merged DataFrame by date and state, and calculate the total transaction

amount and transaction count for each state on each date.

$3.$ Identify the top $5$ states with the highest total transaction amounts for each date and

print the results.

$4.$ Additionally, calculate the daily transaction amount average for each state and print the

top $3$ states with the highest averages over the entire period.

Hint: Use groupby$()$ to aggregate daily transaction data and sort to identify the top states. For this assignment, use the cc$\_$info.csv and transactions.csv files. Your submission should include:

$1.$ The Python script.

$2.$ Sample output.

$3.$ An explanation of your approach and reasoning behind your code.

Each question is worth $10$ points, broken down as follows:

$-$ Python script: $6$ points

$-$ Output: $2$ points

$-$ Code explanation and reasoning: $2$ points

Question $1$: Suspicious Transaction Analysis $(10$ points$)$

Company ABC suspects that transactions that exceed a certain threshold within a short period of time might be fraudulent. Using the transactions DataFrame:

$1.$ Identify all transactions that are greater than $\backslash (\backslash$$ $300\backslash ).$

$2.$ Sort these high$-$value transactions by the credit$\_$card column and then by date.

$3.$ For each credit$\_$card, calculate the time difference in days between consecutive high$-$value transactions.

$4.$ Add a new column called suspicious to indicate whether a transaction is considered suspicious. A transaction is considered suspicious if:

$-$ It occurs within $3$ days of the previous high$-$value transaction for the same credit$\_$card.

$5.$ Print the credit$\_$card, transaction$\_$dollar$\_$amount, date, suspicious columns for all high$-$value transactions.

Hint: Use pd$.$to$\_$datetime $()$ to ensure date format compatibility and diff$()$ to calculate consecutive differences in days.

Question $2$: Outlier Detection in Transaction Amounts $(10$ points$)$

Outlier detection can help identify unusual spending patterns. Using the transactions DataFrame:

$1.$ Calculate the median and interquartile range $($IQR$)$ for the transaction$\_$dollar$\_$amount column.

$2.$ Define transactions as "outliers" if their transaction$\_$dollar$\_$amount is either below Q$1-1\mathrm{.}5*$ IQR or above Q$3+1\mathrm{.}5*$ IQR. For this assignment, use the cc$\_$info.csv and transactions.csv files. Your submission should include:

$1.$ The Python script.

$2.$ Sample output.

$3.$ An explanation of your approach and reasoning behind your code.

Each question is worth $10$ points, broken down as follows:

$-$ Python script: $6$ points

$-$ Output: $2$ points

$-$ Code explanation and reasoning: $2$ points

Question $1$: Suspicious Transaction Analysis $(10$ points$)$

Company ABC suspects that transactions that exceed a certain threshold within a short period of time might be fraudulent. U $3.$ Create a new column n