使用go.Scatter()绘制整个数据框的图表

Question

我的CSV文件如下：

DateTime;S1;S2;S3;S4;S5;S6
2020-07-17 09:57:27.119916;725.9926027110598;730.5730869210306;946.2937510737263;542.341137182406;758.5531610786929;512.2027881299339
2020-07-17 09:57:28.119916;761.1846087077208;984.1009029835216;974.8724733720549;576.8019892357476;751.6553704523698;855.5439493088621
2020-07-17 09:57:29.119916;618.7837289058051;823.9970681226491;594.2841714340789;873.3093170922189;770.0875733375253;681.1715820388949
2020-07-17 09:57:30.119916;515.9456035777555;533.017970929369;639.3409213385498;542.4405737836958;514.4985515824058;650.5229638670448
2020-07-17 09:57:31.119916;589.1350057317254;605.703259361724;602.3181712775759;860.9749699475683;801.7960812507487;562.400896160191
2020-07-17 09:57:32.119916;626.2528314431347;615.7078057434281;643.2023497200336;709.6997180536518;741.365852401098;712.4384053449293
2020-07-17 09:57:33.119916;553.9768845577024;961.7714859567449;519.8207498752649;551.8006708566627;511.7426656331682;849.3428394570542
2020-07-17 09:57:34.119916;994.8208541190293;700.59423301376;569.1853469890981;997.5842090634065;621.2070112896865;848.5079857917269
2020-07-17 09:57:35.119916;502.2301607876932;760.8787524302393;671.2907579865052;669.0718770518221;901.3788876259023;926.077760311429
2020-07-17 09:57:36.119916;578.3978109170034;811.407262562966;822.6244615030105;570.0016494663124;935.0853062150045;689.8800124555897

我使用了这个函数读取了一个csv文件：

df = pd.read_csv(file_name, delimiter = ';', parse_dates=['DateTime']).set_index('DateTime')

df 包含以下内容：

                                    S1          S2  ...          S5          S6
DateTime                                            ...                        
2020-07-17 09:57:27.119916  725.992603  730.573087  ...  758.553161  512.202788
2020-07-17 09:57:28.119916  761.184609  984.100903  ...  751.655370  855.543949
2020-07-17 09:57:29.119916  618.783729  823.997068  ...  770.087573  681.171582
2020-07-17 09:57:30.119916  515.945604  533.017971  ...  514.498552  650.522964
2020-07-17 09:57:31.119916  589.135006  605.703259  ...  801.796081  562.400896
2020-07-17 09:57:32.119916  626.252831  615.707806  ...  741.365852  712.438405
2020-07-17 09:57:33.119916  553.976885  961.771486  ...  511.742666  849.342839
2020-07-17 09:57:34.119916  994.820854  700.594233  ...  621.207011  848.507986
2020-07-17 09:57:35.119916  502.230161  760.878752  ...  901.378888  926.077760
2020-07-17 09:57:36.119916  578.397811  811.407263  ...  935.085306  689.880012

[10 rows x 6 columns]

我想使用plotly.graph_objs.Scatter()来绘制整个数据框中的数据，但是Scatter函数需要x[]和y[]。

如何在X轴上获取DateTime，并且对于每个S1，S2，S...，使用整个Dataframe上的Y值在X时间上进行绘图？

如果不可能，是否有其他使用Plotly库的方法？

- Wes

Answer 1

这是一种方法。你是想将它们绘制到同一图中吗？您可以通过将 lines 替换为 markers 来将数据绘制为点。

import pandas as pd
import plotly.graph_objects as go

file_name = "test.csv"
df = pd.read_csv(file_name, delimiter = ';')

# Create figure
fig = go.Figure()

# Loop df columns and plot columns to the figure
for i in range(1, len(df.columns)):
    col_name = 'S'+ str(i)
    fig.add_trace(go.Scatter(x=df['DateTime'], y=df[col_name],
                        mode='lines', # 'lines' or 'markers'
                        name=col_name))

fig.show()

- Heikura

---

太好了，谢谢。我想问一下如何在不同的图表中完成相同的操作？我的意思是在一个矩阵图中，每个图表都紧随其后。类似于这样：from pandas.plotting import scatter_matrix scatter_matrix(df, alpha=0.2, figsize=(9, 9), diagonal='kde')。 - GSandro_Strongs

你可以使用 Plotly subplots: https://plotly.com/python/subplots/! fig.add_trace(go.Scatter(x=df['DateTime'], y=df[col_name], mode='lines', name=col_name), row=i, col=1) 你只需要定义每行要有多少个图（col 值）。这需要另一个循环或 if 语句。 :) - Heikura

Answer 2

我认为这个线程应该包含一个使用 Plotly.Express 的答案。通过这个包，可以从你传入的 DataFrame 生成一个 Plotly.graph_objs 图形对象。你可以这样做：

import pandas as pd
import plotly.graph_objects as go
import plotly.express as px

file_name = "test.csv"
df = pd.read_csv(file_name, delimiter = ';')

# Create figure
fig = px.line(
    df,
    title = "All curves in one diagram",
)

# Show figure
fig.show()

如果您不打算将对象fig在未来进行修改，则可以使用快捷方式：

import pandas as pd
import plotly.graph_objects as go
import plotly.express as px

file_name = "test.csv"
df = pd.read_csv(file_name, delimiter = ';')

# Create and show figure
px.line(
    df,
    title = "All curves in one diagram",
).show()

如果您想在六个单独的图表中显示曲线，可以使用Plotly.Express以相同的方式进行操作，但是您需要为每个DataFrame列制作一个Figure：

import pandas as pd
import plotly.graph_objects as go
import plotly.express as px

file_name = "test.csv"
df = pd.read_csv(file_name, delimiter = ';')

for column_name in df.columns:  # this will be column_name =
                                # "S1", "S2", ..., "S5"
    # Create and show figure
    px.line(
        df[column_name],
        title = "Curve compiled from " + column_name,
    ).show()

- NerdOnTour