如何使折线图成为一条连续线

debugcn 发表于 Dev

痛风

我试图可视化我的数据，但是当我绘制点时，线条似乎无处不在。

这是数据片段

Date    time_began  time_end    activecalls date_start  date_end
7/3/2020    14:08:47    14:09:30    2   7/3/2020 14:08  7/3/2020 14:09
7/3/2020    14:06:05    14:06:48    4   7/3/2020 14:06  7/3/2020 14:06
7/3/2020    15:11:36    15:12:19    6   7/3/2020 15:11  7/3/2020 15:12
7/3/2020    13:37:52    13:38:35    1   7/3/2020 13:37  7/3/2020 13:38
7/3/2020    14:19:31    14:20:14    3   7/3/2020 14:19  7/3/2020 14:20
7/3/2020    13:58:01    13:58:44    1   7/3/2020 13:58  7/3/2020 13:58
7/3/2020    16:56:32    16:57:15    3   7/3/2020 16:56  7/3/2020 16:57
7/3/2020    16:15:26    16:16:09    6   7/3/2020 16:15  7/3/2020 16:16
7/3/2020    14:35:16    14:35:59    3   7/3/2020 14:35  7/3/2020 14:35
7/3/2020    15:54:48    15:55:31    9   7/3/2020 15:54  7/3/2020 15:55
7/3/2020    16:01:39    16:02:22    3   7/3/2020 16:01  7/3/2020 16:02
7/3/2020    15:52:51    15:53:34    4   7/3/2020 15:52  7/3/2020 15:53

当我运行它时，图表如下所示：看起来像什么

这是我想要的样子：我想要它看起来像什么

舍

您用来绘制数据的代码没有错，只是数据本身与您的期望不符。我在这里做了一些假设，但根据以前的工作，我认为您需要做两件事来纠正此问题

数据中的日期范围

你已经覆盖你数据框限制你的数据只能从包含的信息'7/1/2020 16:08'到'7/4/2020 15:10'这里：

mask = (df['date_start'] > day1) & (df['date_end'] <= day2)
df = df.loc[mask]

I'm not sure if this is intentional just to check the first few days but your expected chart goes up to 2009 so I'd recommend removing these lines.

Active Calls - Try using an aggregation for each day

Looking at the figures in your data snippet and comparing to your expected output the data is quite granular from 2002 - 2009. If you want to aggregate the sum of active calls by day you want to include a groupby() with a pd.Grouper() to specify the frequency of day:

df.groupby(pd.Grouper(key='date_start', freq='D'))['activecalls'].sum()

From here you can simply plot the data adding .plot() which will by default plot a line chart as the index is now your date field (aggregated by day):

df.groupby(pd.Grouper(key='date_start', freq='D'))['activecalls'].sum().plot()

Finally, there's a few inconsistencies in your code it would be good to go through and clean these up:

您不需要两次转换date_start并转换date_end为pd datetime，因此可以删除此事件的第二个实例。
您已经activecalls用另一种创建方法覆盖了该列。确定哪个是正确的，然后删除另一个。

本文收集自互联网，转载请注明来源。

如有侵权，请联系[email protected] 删除。

编辑于2021-04-5

我来说两句

0条评论

登录后参与评论

来自分类Dev

Related 相关文章

文章