带有色条的普雷蒂甘特图

Question

我正在制作这样的甘特图https://plot.ly/python/gantt/#index-然而，通过数值变量，我想用作索引的数值变量是一个大于100的正数。当我使用示例代码时，色条限制为[0100]，导致色条全部为100色。有没有办法在使用示例中的代码创建的绘制甘特图中提升最大值？

我希望颜色与指数的值“成正比”。

以下是示例代码：

import plotly.figure_factory as ff

df = [dict(Task="Job A", Start='2009-01-01', Finish='2009-02-28', Complete=10),
      dict(Task="Job B", Start='2008-12-05', Finish='2009-04-15', Complete=60),
      dict(Task="Job C", Start='2009-02-20', Finish='2009-05-30', Complete=95)]

fig = ff.create_gantt(df, colors='Viridis', index_col='Complete', show_colorbar=True)
fig.show()

需要明确的是：在我的例子中，变量完成可以达到高达700000的值

Answer 1

根据文档，当给定数字索引时，它应该缩放。然而，看看代码，似乎最小/最大值已硬编码为0和100。见代码行285和286。

一种解决方法是在0-100之间缩放索引并手动将颜色栏的标签设置为原始最小/最大值：

import warnings
warnings.filterwarnings('ignore', category=FutureWarning)  # plotly returns a FutureWarning due to using .ix

from sklearn import preprocessing
import plotly.figure_factory as ff

# create a dataframe for easier processing wrt the scaled index
df = pd.DataFrame([dict(Task="Job A", Start='2009-01-01', Finish='2009-02-28', Complete=10),
      dict(Task="Job B", Start='2008-12-05', Finish='2009-04-15', Complete=60),
      dict(Task="Job C", Start='2009-02-20', Finish='2009-05-30', Complete=995)])

# scale index (0-100) and add as new column
min_max_scaler = preprocessing.MinMaxScaler()
index_scaled = min_max_scaler.fit_transform(df.Complete.values.reshape(-1,1))
df['index_scaled'] = index_scaled * 100

# create figure based on scaled index
fig = ff.create_gantt(df, index_col='index_scaled', colors='Viridis', show_colorbar=True)

# set scale of color bar
fig.data[-1]['marker']['cmin'] = df.Complete.min()
fig.data[-1]['marker']['cmax'] = df.Complete.max()
fig.data[-2]['marker']['cmin'] = df.Complete.min()
fig.data[-2]['marker']['cmax'] = df.Complete.max()

# Due to indexing on the scaled value, the tooltip of the bars shows this values instead of its original.
# You could iterate over them and adjust, but you'll need to match the sorting of fig.data to the sorting of the dataframe.
# fig.data[0].name = original_value

fig.show()

请注意，如上面的注释所述，您可能需要注意条形图的工具提示（显示索引值）。这些也可以手动设置，但您需要将fig.data的顺序与原始数据框的顺序相匹配。