Matplotlib散点图X轴刻度标签问题及Pandas条件绘图优化咨询

Hey there! Let's tackle your two questions about plotting poker hand data—first fixing that wonky X-axis tick issue, then streamlining your conditional plotting workflow.

一、X轴刻度标签异常的修复方案 #

The root cause of your tick problem is likely a mismatch between your manual tick settings and the actual range of your RFI data. You set plt.xticks(x) with x = np.arange(0,1,0.1), but if your RFI values are percentages (like 0-100 instead of 0-1 decimals), this will throw off the scale entirely.

Step 1: Verify your data range first #

Start by checking what your RFI values actually look like to confirm the right scale:

print(preflop_copy['RFI'].describe())

Step 2: Adjust ticks to match your data #

• If RFI is a percentage (0-100):

  # Set ticks from 0 to 100 in 10-point increments
  plt.xticks(np.arange(0, 101, 10))
  

• If RFI is a decimal (0-1):

  # Include the 1.0 endpoint to avoid cutting off the upper range
  plt.xticks(np.arange(0, 1.01, 0.1))
  

• For zero hassle, skip manual ticks entirely: Let Matplotlib auto-generate ticks that fit your data by removing the plt.xticks(x) line.

Bonus: Fix overlapping labels #

If your tick labels are crammed together, add these lines to clean up the layout:

plt.xticks(rotation=45)
plt.tight_layout() # Automatically adjusts padding to prevent label cutoff

二、无需创建新列表的Pandas条件绘图方法 #

Absolutely no need to make a separate filtered DataFrame! Pandas lets you combine filtering and plotting in one step using boolean indexing or the query() method for cleaner code.

Method 1: Boolean indexing (direct and straightforward) #

plt.figure(figsize=(8,8))
# Filter and plot in a single line
preflop_copy[preflop_copy['Hands'] > 50].plot.scatter(x="RFI", y="BB_100", ax=plt.gca())

# Add your labels and optimize layout
plt.xlabel("RFI", fontsize=16)
plt.ylabel("BB/100", fontsize=16)
plt.tight_layout()
plt.show()

Method 2: query() method (more readable for complex conditions) #

Use SQL-like syntax for filtering—great if you ever need to add multiple conditions later:

plt.figure(figsize=(8,8))
# Filter with query() and plot
preflop_copy.query("Hands > 50").plot.scatter(x="RFI", y="BB_100", ax=plt.gca())

plt.xlabel("RFI", fontsize=16)
plt.ylabel("BB/100", fontsize=16)
plt.tight_layout()
plt.show()

Why this works better:

• Saves memory by avoiding intermediate DataFrames (especially useful with large datasets)
• Keeps your code logic cohesive—from data filter to plot in one flow

Full optimized code putting it all together #

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# Data loading and preprocessing
preflop = pd.read_csv("all_player_preflop_report_tourney.csv", thousands=',')
preflop['Hands'] = preflop['Hands'].astype(int)

# Clean up column names with a mapping (easier to read than long chained renames)
column_renames = {
    'BB/100':'BB_100',
    'Raise First':'RFI',
    'WTSD %': 'WTSD', 
    'All-In Adj BB/100':'adj_BB_100',
    'Avg PF All-In Equity':'pf_all_in',
    'CC 2Bet PF':'cc_2bet',
    '3Bet PF':'3bet',
    '2Bet PF & Call 3Bet':'2Bet_call_3Bet',
    'Raise & 4Bet+ PF':'rfi_and_4bet+',
    '2Bet PF & Fold':'2bet_and_fold',
    '5Bet+ PF':'5bet+',
    '3Bet PF & Fold':'3bet_and_fold',
    'Call Any PFR':'call_any_pfr',
    'Call Steal':'call_steal', 
    'Call vs BTN Open':'call_btn_open',
    'CC 3Bet+ PF':'cc_3bet+',
    'Limp Behind':'limp_behind',
    'Raise Limpers':'raise_limpers'
}
preflop = preflop.rename(columns=column_renames).set_index('Player')
preflop_copy = preflop.copy()

# Conditional plotting without intermediate DataFrames
plt.figure(figsize=(8,8))
preflop_copy.query("Hands > 50").plot.scatter(x="RFI", y="BB_100", ax=plt.gca())

# Final chart polish
plt.xlabel("RFI", fontsize=16)
plt.ylabel("BB/100", fontsize=16)
# Uncomment below if you want manual ticks (adjust range to match your RFI data)
# plt.xticks(np.arange(0, 101, 10))
plt.tight_layout()
plt.show()

内容的提问来源于stack exchange，提问作者jonstoeber