pandas 在 Python 中绘制直方图的时间序列
声明：本页面是StackOverFlow热门问题的中英对照翻译，遵循CC BY-SA 4.0协议，如果您需要使用它，必须同样遵循CC BY-SA许可，注明原文地址和作者信息，同时你必须将它归于原作者（不是我）：StackOverFlow
原文地址: http://stackoverflow.com/questions/17050202/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me):
StackOverFlow
Plot timeseries of histograms in Python
提问by abudis
I'm trying to plot a time-series of histograms in Python. There has been a similar question about this, but in R. So, basically, I need the same thing, but I'm really bad in R. There are usually 48 values per day in my dataset. Where - 9999 represents missing data. Here'sthe sample of the data.
我正在尝试在 Python 中绘制直方图的时间序列。有一个类似的问题，但在 R 中。所以，基本上，我需要同样的东西，但我在 R 方面真的很糟糕。我的数据集中每天通常有 48 个值。其中 - 9999 表示缺失数据。 这是数据的示例。
I started with reading in the data and constructing a pandasDataFrame.
我首先读入数据并构建一个pandasDataFrame.
import pandas as pd
df = pd.read_csv('sample.csv', parse_dates=True, index_col=0, na_values='-9999')
print df
<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 336 entries, 2008-07-25 14:00:00 to 2008-08-01 13:30:00
Data columns (total 1 columns):
159.487691046 330 non-null values
dtypes: float64(1)
Now I can group the data by day:
现在我可以按天对数据进行分组：
daily = df.groupby(lambda x: x.date())
But then I'm stuck. I don't know how to use this with matplotlibto get my timeseries of histograms. Any help appreciated, not necessarily using pandas.
但后来我被卡住了。我不知道如何使用它matplotlib来获取我的直方图时间序列。任何帮助表示赞赏，不一定使用pandas.
采纳答案by Dan Allan
Make a histogram and use matplotlib's pcolor.
制作直方图并使用 matplotlib 的pcolor.
We need to bin the groups uniformly, so we make bins manually based on the range of your sample data.
我们需要对组进行统一分箱，因此我们会根据您的样本数据范围手动进行分箱。
In [26]: bins = np.linspace(0, 360, 10)
Apply histogramto each group.
适用histogram于每个组。
In [27]: f = lambda x: Series(np.histogram(x, bins=bins)[0], index=bins[:-1])
In [28]: df1 = daily.apply(f)
In [29]: df1
Out[29]:
0 40 80 120 160 200 240 280 320
2008-07-25 0 0 0 3 18 0 0 0 0
2008-07-26 2 0 0 0 17 6 13 1 8
2008-07-27 4 3 10 0 0 0 0 0 31
2008-07-28 0 7 15 0 0 0 0 6 20
2008-07-29 0 0 0 0 0 0 20 26 0
2008-07-30 10 1 0 0 0 0 1 25 9
2008-07-31 30 4 1 0 0 0 0 0 12
2008-08-01 0 0 0 0 0 0 0 14 14
Following your linked example in R, the horizontal axis should be dates, and the vertical axis should be the range of bins. The histogram values are a "heat map."
按照 R 中的链接示例，横轴应为日期，纵轴应为 bin 的范围。直方图值是一个“热图”。
In [30]: pcolor(df1.T)
Out[30]: <matplotlib.collections.PolyCollection at 0xbb60e2c>
It remains to label the axes. This answershould be of some help.
剩下来标记轴。这个答案应该会有所帮助。
