Python绘制频谱图遇TypeError: 'float' object is not callable错误求助

原代码 #

import numpy as np
import matplotlib.pyplot as plt


sampling_frequency_48kHz = 48000
t_48kHz = np.linspace(0, 10, sampling_frequency_48kHz * 10)
signal_48kHz = np.sin(2 * np.pi * 5 * t_48kHz)

sampling_frequency_8kHz = 8000
t_8kHz = np.linspace(0, 10, sampling_frequency_8kHz * 10)
signal_8kHz = np.sin(2 * np.pi * 5 * t_8kHz)

window_size = 0.03 
overlap = 0 
nfft = 2048 

plt.figure()
plt.specgram(signal_48kHz, Fs=sampling_frequency_48kHz, window=window_size, noverlap=overlap, NFFT=nfft)
plt.title('Spektrogram pro 48 kHz')

nfft = 512 

plt.figure()
plt.specgram(signal_8kHz, Fs=sampling_frequency_8kHz, window=window_size, noverlap=overlap, NFFT=nfft)
plt.title('Spektrogram pro 8 kHz')

plt.show()

错误信息 #

Traceback (most recent call last):
  File "c:\Users\OneDrive\Dokumenty\4_7.py", line 18, in <module>
    plt.specgram(signal_48kHz, Fs=sampling_frequency_48kHz, window=window_size, noverlap=overlap, NFFT=nfft)
  File "C:\Users\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\matplotlib\pyplot.py", line 2818, in specgram
    __ret = gca().specgram(
  File "C:\Users\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\matplotlib\__init__.py", line 1423, in inner
    return func(ax, *map(sanitize_sequence, args), **kwargs)
  File "C:\Users\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\matplotlib\axes\_axes.py", line 7709, in specgram
    spec, freqs, t = mlab.specgram(x=x, NFFT=NFFT, Fs=Fs,
  File "C:\Users\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\matplotlib\mlab.py", line 780, in specgram
    spec, freqs, t = _spectral_helper(x=x, y=None, NFFT=NFFT, Fs=Fs,
  File "C:\Users\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\matplotlib\mlab.py", line 377, in _spectral_helper
    window = window(np.ones(NFFT, x.dtype))
TypeError: 'float' object is not callable

错误原因 #

错误核心是specgram函数的window参数使用错误：

• 你传入的window_size是浮点型数值（0.03，代表时间长度），但该参数要求的是可调用的窗口函数（比如np.hanning）、窗口数组，或者字符串形式的窗口名称（比如'hann'）。
• 若要设置窗口的时间长度，需要将其转换为对应采样率下的样本数，再通过nperseg参数传入，该参数用于指定每个分析段的样本数量。

修正后的代码 #

import numpy as np
import matplotlib.pyplot as plt


sampling_frequency_48kHz = 48000
t_48kHz = np.linspace(0, 10, sampling_frequency_48kHz * 10)
signal_48kHz = np.sin(2 * np.pi * 5 * t_48kHz)

sampling_frequency_8kHz = 8000
t_8kHz = np.linspace(0, 10, sampling_frequency_8kHz * 10)
signal_8kHz = np.sin(2 * np.pi * 5 * t_8kHz)

window_size = 0.03 
overlap = 0 
nfft = 2048 

# 计算对应采样率下的窗口样本数
window_samples_48kHz = int(window_size * sampling_frequency_48kHz)

plt.figure()
# 使用nperseg指定窗口样本数，window参数指定窗口类型（可选，默认汉宁窗）
plt.specgram(signal_48kHz, Fs=sampling_frequency_48kHz, nperseg=window_samples_48kHz, noverlap=overlap, NFFT=nfft, window='hann')
plt.title('Spektrogram pro 48 kHz')

nfft = 512 
window_samples_8kHz = int(window_size * sampling_frequency_8kHz)

plt.figure()
plt.specgram(signal_8kHz, Fs=sampling_frequency_8kHz, nperseg=window_samples_8kHz, noverlap=overlap, NFFT=nfft, window='hann')
plt.title('Spektrogram pro 8 kHz')

plt.show()

说明 #

• nperseg参数控制每个分析窗口的样本数量，通过时间长度 × 采样率计算得到，符合你设置window_size=0.03秒的需求。
• window参数可指定窗口类型，比如'hann'（汉宁窗）、'hamming'（汉明窗）等，默认也是汉宁窗，可根据需求调整。

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