Multiple boxplot from multiple dataframes-CodePudding

I have 4 dataframes

df1 = pd.DataFrame({'ID': [0, 0, 0, 0, 0, 0],
                    'value': [3.0, 3.5, 4.5, NaN, 7.0, 8.1]})

df2 = pd.DataFrame({'ID': [1, 1, 1, 1, 1, 1],
                    'value': [9.4, NaN, 4.5, 2.4, 4.0, 3.9]})

df3 = pd.DataFrame({'ID': [2, 2, 2],
                    'value': [1.0, 3.9, 4.1]})

df4 = pd.DataFrame({'ID': [3, 3, 3, 3],
                    'value': [NaN, NaN, 5.8, 3.0]})

I want to make a boxplot with values in the column value in each of the dataframe. I did the following

fig, ax2 = plt.subplots()
vec = [df1['value'].values,df2['value'].values,df3['value'].values,df4['value'].values]
labels = ['ID_0','ID_1', 'ID_2', 'ID_3']
ax2.boxplot(vec, labels = labels)
ax2.set_title('Values')
plt.show()

But it doesn't work and throws me an empty plot. Is there a better way to do this?

CodePudding user response：

To identify NaN, you need to use np.nan (use import numpy as np if required). Also, you need to dropna() before plotting. Making the changes...

df1 = pd.DataFrame({'ID': [0, 0, 0, 0, 0, 0], 'value': [3.0, 3.5, 4.5, np.nan, 7.0, 8.1]}).dropna()
df2 = pd.DataFrame({'ID': [1, 1, 1, 1, 1, 1], 'value': [9.4, np.nan, 4.5, 2.4, 4.0, 3.9]}).dropna()
df3 = pd.DataFrame({'ID': [2, 2, 2], 'value': [1.0, 3.9, 4.1]}).dropna()
df4 = pd.DataFrame({'ID': [3, 3, 3, 3],'value': [np.nan, np.nan, 5.8, 3.0]}).dropna()
fig, ax2 = plt.subplots()
vec = [df1['value'].values,df2['value'].values,df3['value'].values,df4['value'].values]
labels = ['ID_0','ID_1', 'ID_2', 'ID_3']
ax2.boxplot(vec, labels = labels)
ax2.set_title('Values')
plt.show()

gives you...