Python practice: Exploratory Data Analysis (EDA) Workflow

In professional data science, what are the five foundational steps of the Exploratory Data Analysis (EDA) workflow?

Load "orders_messy.csv". Inspect the table shape and print the number of rows (df.shape[0]):

Load "orders_messy.csv". Count missing values per column with df.isnull().sum(). Print the count of missing values in the "Units" column (int(nulls["Units"])):

Load "sales.csv". Compute Q1 (sales["Revenue"].quantile(0.25)), Q3 (sales["Revenue"].quantile(0.75)), and IQR = Q3 - Q1. Filter outliers where (sales["Revenue"] < Q1 - 1.5 * IQR) | (sales["Revenue"] > Q3 + 1.5 * IQR). Print outlier count (len(outliers)):

Load "sales.csv". Calculate the skewness of the "Revenue" column using sales["Revenue"].skew(). Print the skewness rounded to 2 decimal places (round(sales["Revenue"].skew(), 2)):

Load "sales.csv". Calculate the percentage share of orders from each region using sales["Region"].value_counts(normalize=True) * 100. Print the North region percentage rounded to 1 decimal place (round(shares["North"], 1)):

Load "sales.csv". Compute the median Revenue by Region using sales.groupby("Region")["Revenue"].median(). Print the median revenue for "South" as an integer:

Load "sales.csv". Construct a summary dictionary scorecard = {"total_rev": int(sales["Revenue"].sum()), "total_orders": len(sales), "avg_units": round(sales["Units"].mean(), 1)}. Print scorecard["total_rev"]: