如何使用Pandas Dataframe在子图直方图中设置多个平均值

Question

If I use the pd.DataFrame.hist() func of a DataFrame with multiple headers, then python will plot multiple histograms.

I wanted to plot the mean with the plt.axvline function with the mean of the dataframe. This is not working

I have already tried it with only one header of a dataframe and it worked.

def plot_HistOfDailyReturn(p_df, p_bins=10):

    """plots the histogram of the daily returns"""                          
    df1 = pd.DataFrame(p_df['HCP'])                           
    df1.hist(bins=p_bins)
    plt.axvline(df1['HCP'].mean(), color='w', linestyle='dashed', linewidth=2)
    plt.show()

How can I now apply this to several ones? But I don't want want them to be each in a seperate plot.

Answer 1

DataFrame.hist() returns a matplotlib.AxesSubplot or numpy.ndarray of them. If you have multiple columns, then most probably it'll return a 2D numpy.ndarray of matplotlib.AxesSubplot . You can use those AxesSubplot's in order to draw lines in each subplot.

Demo:

In [199]: data = pd.DataFrame(np.random.randint(100, size=(1000,4)), columns=list("abcd"))

In [200]: data.shape
Out[200]: (1000, 4)

In [201]: (a, b), (c, d) = data.hist(bins=100, alpha=0.8, figsize=(12, 8))

In [202]: a.axvline(data["a"].mean(), color='orange', linestyle='dashed', linewidth=2)
Out[202]: <matplotlib.lines.Line2D at 0x149c220a4e0>

In [203]: b.axvline(data["b"].mean(), color='orange', linestyle='dashed', linewidth=2)
Out[203]: <matplotlib.lines.Line2D at 0x149c220aa58>

In [204]: c.axvline(data["c"].mean(), color='orange', linestyle='dashed', linewidth=2)
Out[204]: <matplotlib.lines.Line2D at 0x149c098f5f8>

In [205]: d.axvline(data["d"].mean(), color='orange', linestyle='dashed', linewidth=2)
Out[205]: <matplotlib.lines.Line2D at 0x149c222d550>