Python：利用循环替换DataFrame列中的指定0值

Problem Breakdown #

Your current code has a few key issues that are preventing the correct replacement:

1. When using inplace=True, the replace() method returns None—assigning this back to result will overwrite your DataFrame with a None value, which breaks your data.
2. You're targeting the string '0', but your DataFrame contains numeric 0 values, so the match won't work.
3. You haven't tied the loop logic to the specific rows that should get each replacement value from your contas list.

Solution #

Based on your description, it sounds like you want to replace 0 values in the Contas Resultado column with the corresponding financial metric from your contas list (e.g., replace 0 in rows associated with "Faturamento" with "Faturamento"). Here's the corrected approach:

contas = ["Margem Bruta","Faturamento", "Resultado Bruto", "Resultado Líquido", "Despesas Operacionais", "Despesas Vendas"]
results = []

# Assume your initial query result is stored in this DataFrame
df = ... # Your query-generated DataFrame

for conta in contas:
    # Create a mask to target rows where:
    # 1. The Contas column matches the current metric
    # 2. Contas Resultado is 0
    target_rows = (df['Contas'] == conta) & (df['Contas Resultado'] == 0)
    # Replace the 0 values with the current metric name
    df.loc[target_rows, 'Contas Resultado'] = conta

# Add the processed DataFrame to your results list
results.append(df)

Key Notes #

• Using .loc for conditional assignment is the safest way to modify DataFrames in Pandas, avoiding chained assignment warnings.
• We're matching numeric 0 instead of the string '0' to ensure we target the correct rows.
• The loop ensures each metric from contas only affects rows associated with that metric, keeping your replacements aligned with your data structure.

内容的提问来源于stack exchange，提问作者Cesar