使用Seaborn时出现AttributeError：float对象无shape属性的问题

Let's break down why you're running into this error and fix your simulated tips dataset code properly.

What's Causing the Error? #

The AttributeError pops up because Seaborn expects valid pandas Series or numpy arrays for plotting, but your DataFrame likely has columns with messed-up data types—like categorical columns such as time/sex/smoker accidentally getting converted to float.

Looking at your code snippet, the main issues are:

• You only initialize the DataFrame for the first combination (t=0, s=0, sm=0) and never properly add data from other category combinations.
• Using df.loc[i] in nested loops is error-prone; it’s easy to mess up row indexing or accidentally assign the wrong data type to columns.

Corrected Code to Simulate the Tips Dataset #

Instead of clunky nested loops with manual row assignments, we can generate all your data in a clean, scalable way:

import numpy as np
import pandas as pd

# Define all possible values for categorical variables
time_options = ['day', 'night']
sex_options = ['female', 'male']
smoker_options = ['yes', 'no']

# Generate all combinations of the categorical variables
category_combinations = [(t, s, sm) for t in time_options 
                         for s in sex_options 
                         for sm in smoker_options]

# Collect all rows of data
all_data = []
for time_val, sex_val, smoker_val in category_combinations:
    # Generate 10 random total_bill values (range 0-10)
    total_bills = np.random.rand(10) * 10
    # Append each bill with its corresponding category labels
    for bill in total_bills:
        all_data.append([bill, time_val, sex_val, smoker_val])

# Create the DataFrame in one go (far less error-prone!)
df = pd.DataFrame(all_data, columns=["total_bill", "time", "sex", "smoker"])

# Verify data types to ensure no unexpected floats in categorical columns
print(df.dtypes)

Why This Works #

• We generate all category combinations upfront, then build a list of rows to avoid incremental DataFrame modifications (which are slow and easy to botch).
• Categorical columns (time, sex, smoker) are explicitly stored as strings, so Seaborn can correctly interpret them as categorical data without hitting type errors.
• You can easily adjust the number of samples per combination (just change 10 to any number) without rewriting the entire loop logic.

Quick Fix If You Still Run Into Issues #

If you ever see this error again, first check your DataFrame's data types with df.dtypes. If a categorical column shows up as float, convert it back to string with:

df['time'] = df['time'].astype(str)
df['sex'] = df['sex'].astype(str)
df['smoker'] = df['smoker'].astype(str)

内容的提问来源于stack exchange，提问作者lanadaquenada