CSV数据加载至Bokeh失败，数据库加载正常，请求技术支持

Hey there! Let's work through this Bokeh loading issue step by step. Looking at your code, I spot a few key problems that are keeping your CSV-loaded data from rendering correctly:

Key Issues in Your Code #

• Index vs. Column Mix-Up: You set index_col='Year' which makes Year the DataFrame's index instead of a regular column. This breaks two things:
  1. The groupby('Year') call fails because Year isn't a column anymore.
  2. When creating ColumnDataSource(df), the index isn't included by default, so the x=index in your circle() call can't find that field.
• Invalid x_range Value: You passed a groupby object to x_range, but Bokeh needs a list of categorical values (like your year labels) or a numeric range to render the axis properly.
• Redundant Code: df = pd.DataFrame(df) is unnecessary—pd.read_csv() already returns a DataFrame.

Fixed Code Example #

Here's a revised version that addresses all these issues:

import pandas as pd
from bokeh.plotting import figure, show, output_notebook
from bokeh.models import ColumnDataSource

# Load CSV, keep Year as a column (no need to set it as index here)
df = pd.read_csv('unemployment1948.csv', delimiter=',')

# Filter to the rows you want (35th row onwards)
filtered_df = df[35:]

# Create the data source with our filtered data
source = ColumnDataSource(filtered_df)

# Set x_range to the unique years from our filtered dataset
p = figure(plot_height=300, plot_width=900,
           x_range=filtered_df['Year'].unique().tolist(),
           title='Unemployment over the years',
           x_axis_label='Year', y_axis_label='Annual')

# Use the column names from the data source for x/y values
p.circle(x='Year', y='Annual', color='#35B778', source=source)

output_notebook()
show(p)

If You Need Year as an Index #

If you rely on Year being the index for other operations, just reset it to a column before creating the data source:

df = pd.read_csv('unemployment1948.csv', delimiter=',', index_col='Year')
# Convert index back to a column
df = df.reset_index()
filtered_df = df[35:]

Quick Final Check #

Double-check your CSV file to make sure the column names Year and Annual are spelled correctly (no extra spaces or typos)—that's a super common gotcha too!

内容的提问来源于stack exchange，提问作者ReinholdN