R语言：如何基于时间创建符号变化值的直方图？

Let’s break this into two straightforward parts: processing your data to calculate how long each positive/negative state lasts, then visualizing those durations with a histogram. I’ll use Python with pandas and matplotlib—tools that are standard for this kind of data work.

Step 1: Process Your Data to Get State Durations #

First, let’s assume your dataset has two columns: timestamp (datetime format) and value (the numerical values you’re tracking). Here’s how to compute the duration of each consecutive sign segment:

1. Import the necessary libraries:

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

2. Load and clean your data:
   Make sure your timestamp column is parsed as datetime (if it isn’t already):

df = pd.read_csv("your_data.csv")
df['timestamp'] = pd.to_datetime(df['timestamp'])

3. Label each value’s sign:
   Use np.sign() to tag values as positive (1), negative (-1), or zero (0). If you want to treat zeros as part of the previous non-zero state, add the optional fill step:

df['sign'] = np.sign(df['value'])
# Optional: Fill zeros with the last non-zero sign
df['sign'] = df['sign'].replace(0, method='ffill')

4. Group consecutive sign segments:
   Create a unique ID for each stretch of the same sign—this lets us calculate how long each segment lasts:

# Flag rows where the sign changes from the previous row
df['sign_changed'] = df['sign'] != df['sign'].shift(1)
# Assign a unique ID to each consecutive sign segment
df['segment_id'] = df['sign_changed'].cumsum()

5. Calculate duration for each segment:
   Group by segment ID and sign, then compute the time difference between the first and last timestamp in each group:

duration_df = df.groupby(['segment_id', 'sign']).agg(
    start_time=('timestamp', 'first'),
    end_time=('timestamp', 'last')
).reset_index()

# Convert duration to hours (swap to /60 for minutes, or remove division for seconds)
duration_df['duration_hours'] = (duration_df['end_time'] - duration_df['start_time']).dt.total_seconds() / 3600

Now you have a dataframe with every sign segment’s duration in hours.

Step 2: Build the Histogram #

You can visualize this data in two useful ways—either combine all durations, or split them by sign to compare positive vs negative states.

Option 1: Histogram of All State Durations #

plt.figure(figsize=(10, 6))
plt.hist(duration_df['duration_hours'], bins=15, edgecolor='black')
plt.title('Distribution of Sign State Durations')
plt.xlabel('Duration (Hours)')
plt.ylabel('Number of Segments')
plt.grid(axis='y', alpha=0.75)
plt.show()

Option 2: Side-by-Side Histograms for Positive vs Negative States #

positive_durations = duration_df[duration_df['sign'] == 1]['duration_hours']
negative_durations = duration_df[duration_df['sign'] == -1]['duration_hours']

plt.figure(figsize=(12, 6))
plt.hist([positive_durations, negative_durations], bins=15, label=['Positive', 'Negative'], edgecolor='black')
plt.title('Positive vs Negative State Duration Distribution')
plt.xlabel('Duration (Hours)')
plt.ylabel('Number of Segments')
plt.legend()
plt.grid(axis='y', alpha=0.75)
plt.show()

Quick Tips #

• Handling Zeros: If zeros should be their own state, skip the ffill step and add a label for 0 in your histogram.
• Adjust Time Units: Swap the division in the duration calculation to fit your needs (e.g., /60 for minutes).
• Tweak Bins: Play with the bins parameter in plt.hist() to make the histogram’s detail level match your data.

That’s it—you’ll have a clear histogram showing how long each sign state persisted in your dataset!

内容的提问来源于stack exchange，提问作者Ch Gabri