python - 如何用线连接非纳米值？

Question

我有一个数据框，我想将其绘制在三个子图中。

索引是一个日期时间，因此 x 轴在子图中对齐非常重要。

我的框架的一列仅包含一年中第一天的数据。其余的都是nan。

这是一个示例。

           Cases    Coverage    Measles
2011-01-31  31.0    93.1    29.964546402502616
2011-02-28                  43.68275418275418
2011-03-31                  43.28153153153155
2011-04-30                  37.25000000000002
2011-05-31                  75.5229357798165
2011-06-30                  48.727064220183486
2011-07-31                  30.467948717948733
2011-08-31                  31.286418094278353
2011-09-30                 36.43205930461145
2011-10-31                  31.816645136232765
2011-11-30                  29.399802310147145
2011-12-31                  31.34712643678163
2012-01-31  8.0 93.3        0.80327868852459

我想用虚线连接非 nan Cases 数据，但由于剩余月份的值是 nan，因此不会绘制线条。

知道如何解决这个问题吗？

Answer 1

from io import StringIO
import pandas as pd

txt = """date       Cases    Coverage    Measles
2011-01-31  31.0    93.1    29.964546402502616
2011-02-28  nan     nan     3.68275418275418
2011-03-31  nan     nan     43.28153153153155
2011-04-30  nan     nan     37.25000000000002
2011-05-31  nan     nan     75.5229357798165
2011-06-30  nan     nan     48.727064220183486
2011-07-31  nan     nan     30.467948717948733
2011-08-31  nan     nan     31.286418094278353
2011-09-30  nan     nan     36.43205930461145
2011-10-31  nan     nan     31.816645136232765
2011-11-30  nan     nan     29.399802310147145
2011-12-31  nan     nan     31.34712643678163
2012-01-31  8.0     93.     0.80327868852459
"""

df = pd.read_csv(StringIO(txt), delim_whitespace=True, index_col=0, parse_dates=[0])

# subset df with only columns that have nulls
dfna = df.loc[:, df.isnull().any()]

# plot the first df
ax = df.plot(marker='o')

# then fill in nulls with interpolate
# plot the result with dashed lines
dfna.interpolate('index').plot(ls='--', ax=ax)