x轴范围内的熊猫箱图

Question

I have the foll. dataframe in pandas:

value   combined_delta
100 7
100 45
100 49
100 12
100 71
94.09   21
91.42   45
88.7    36
87.26   34
77.61   55
76.3    28
73.81   89
71  80
69.5    27
67.45   12
66.96   127
66.18   43
54.48   68
54.15   23
53.29   48
53.29   49
53.25   302
51.99   24
51.99   116
50.73   22
49.7    101  
31.05   107
31  63
30.19   116
30.12   58
29.38   31
29.18   8
29  104
28.6    61
28.6    63
28.56   60
28.11   35
27.36   50
27.32   63
26.87   103
26.87   257
26.42   55
26.35   85
26.1    27
25.79   21
25.79   66
25.66   77
25.53   9
25.46   92
25.46   248
24.67   15
24.6    93
24.39   5
24.01   28
24.01   82
23.86   19
23.18   133
22.71   41
22.62   37
21.81   43
21.52   34
21.43   35
21.23   40
21.23   25
20  75
19.98   31
19.98   44
19.84   12
19.82   62
18.83   26
18.71   202
18.02   7
18  28
17.99   39
17.75   40
17.68   81
17.67   16
17.55   54
17.25   13
16.63   19
12.14   22
12.01   24
12  59
11.95   49
11.54   39

How can I make a boxplot where the x-axis shows ranges for value : 0-20, 20-40, 40-60,60-80,80-100 and the y-axis shows combined_delta values?

I can use seaborn like this:

ax = sns.boxplot(x="value", y="combined_delta", data=df)

However, this will not create the ranges from the x-axis the way I want

Answer 1

Use pd.cut to create the groups you want and plot combined_delta against those.

df['value_group'] = pd.cut(df['value'], bins=range(0, 101, 20))
ax = sns.boxplot(x="value_group", y="combined_delta", data=df)

Answer 2

cut = pd.cut(df.value, [0, 20, 40, 60, 80, 100])

boxdf = df.groupby(cut) \
    .apply(lambda df: df.combined_delta.reset_index(drop=True)) \
    .unstack(0)

sns.boxplot(data=boxdf);