Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Plotly.Make_Subplots function does not support transforms property #1172

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
ekn1v opened this issue Sep 10, 2018 · 4 comments
Closed

Plotly.Make_Subplots function does not support transforms property #1172

ekn1v opened this issue Sep 10, 2018 · 4 comments

Comments

@ekn1v
Copy link

ekn1v commented Sep 10, 2018

Hi,

I’m trying to do the link below, but in one column and multiple rows.
https://plot.ly/~empet/14352/defining-subplots-with-two-y-axes-in-eac/#/

My challenge is that my data needs to use transforms because I want to aggregate it and filter it. I have been able to make it in one chart.
https://discourse-cdn-sjc2.com/standard17/uploads/plot/original/2X/1/1069c5ebb154a09505ac0ea21f8dc3e5f09c1f02.png

But when I try to do multiple charts using the plotly.make_subplots function, I get an error:
“ValueError: Invalid property specified for object of type plotly.graph_objs.Bar: ‘transforms’”

########################
#CODE
########################

import pandas as pd
import numpy as np
import scipy as sp
import seaborn
import plotly
import plotly.plotly as py
import plotly.figure_factory as ff
import plotly.graph_objs as go
import pandas as pd
import matplotlib.pyplot as plt
import csv
import plotly.offline as off
from plotly.offline import init_notebook_mode, iplot, plot
from plotly import tools

plotly.tools.set_credentials_file(username='XXX', api_key='XXX')

off.init_notebook_mode(connected=False)

dp = pd.read_csv("FILEPATH")

dp['Week_Ending'] = pd.to_datetime(dp['Week_Ending'])
dp['Event_Description'] = dp['Event_Description'].astype(str)
dp[dp == 0] = np.nan
dp.sort_values(by=['Week_Ending','Customer'], ascending=True, na_position='first')

mylist = dp['Customer']
myset = set(mylist)
myNewList = list(myset)
aggs = myNewList
agg = []
agg_func = []
for i in range(0, len(aggs)):
    agg = dict(
        args=['transforms[0].value', aggs[i]],
        label=aggs[i],
        method='restyle'
        )
    agg_func.append(agg)

mylist2 = dp['Product_Type']
myset2 = set(mylist2)
myNewList2 = list(myset2)
aggs2 = myNewList2
agg2 = []
agg_func2 = []
for i in range(0, len(aggs2)):
    agg2 = dict(
        args=['transforms[1].value', aggs2[i]],
        label=aggs2[i],
        method='restyle'
    )
    agg_func2.append(agg2)


listTraces = {}
traceCount = 1
xCount = 1
yCount = 1
#for x in range(1,10):
for retail in aggs:
    newTrace = dict(
    #        listTraces["trace{0}".format(i)]=
    #        dict(
                x= dp['Week_Ending'], 
                y= dp['Gross_Sales'], 
                name= "Gross_Sales", 
                text= dp['Event_Description'], 
                type= "bar",
#                        xaxis= "x" + str(xCount),
#                        yaxis= "y" + str(yCount), 
                transforms =
                    [
                    dict(
                        type = 'filter',
                        target = dp['Customer'],
                        orientation = '=',
                        value =  retail
                        ),
                    dict(
                        type = 'filter',
                        target = dp['Product_Type'],
                        orientation = '=',
                        value =  'HERO'
                        ),
                    dict(
                        type = "aggregate",
                        groups = dp['Week_Ending'],
                        aggregations = 
                            [
                                dict(target = "y", func = "sum", enabled = True),
                            ]
                        )
                    ]
                )            
#            )
    listTraces["trace{0}".format(traceCount)]=newTrace
    traceCount = traceCount+1
    yCount = yCount+1
    newTrace = dict(
#        listTraces["trace{0}".format(i)]=
#            dict(
                x= dp['Week_Ending'], 
                y= dp['ASP'], 
                name= "ASP", 
                text= dp['Event_Description'], 
                type= "scatter", 
                mode= "lines+markers+text",
#               xaxis= "x" + str(xCount),
#               yaxis= "y" + str(yCount), 
                connectgaps= True,
        #        zeroline= True,
        #        categorryarray= "column",
        #        categoryorder= "category ascending",
        #        mirror= True,
                transforms= 
                    [
                    dict(
                        type = 'filter',
                        target = dp['Customer'],
                        orientation = '=',
                        value =  retail
                        ),
                    dict(
                        type = 'filter',
                        target = dp['Product_Type'],
                        orientation = '=',
                        value =  'HERO'
                        ),
                    dict(
                        type= "aggregate",
                        groups= dp['Week_Ending'],
                        aggregations = 
                            [
                                dict(target = "y", func = "avg", enabled = True),
                            ]
                        )
                    ]
                )    
#            )
    listTraces["trace{0}".format(traceCount)]=newTrace
    traceCount = traceCount+1
    xCount = xCount+1

layout = {"autosize": True, "hovermode": "closest", "showlegend": True}

rows=len(aggs)
fig = plotly.tools.make_subplots(rows=rows, cols=1, vertical_spacing=0.8/rows)

for i in range(1, len(listTraces) + 1):
    if i%2==1:
        fig.append_trace(listTraces["trace"+str(i)], i, 1)                              
    else:
        fig.append_trace(listTraces["trace"+str(i)], i-1, 1)                              

ySpace = 0.2 
xStart = 0
xGap = 1
yStart = 0
yGap = 0.3
tNum = 1
for i in range(1, len(aggs) + 1):
    layout["xaxis"+str(i)]= dict(
        autorange= True, 
        domain= [xStart, xStart+xGap], 
        range= ["2018-05-08 12:00", "2018-08-07 12:00"], 
        title= "Week_Ending", 
        type= "date")
    layout["yaxis"+str(tNum)]= dict(
        autorange= True, 
        title= "Gross_Sales", 
        domain= [yStart, yStart+yGap], 
        type= "linear")
    tNum = tNum + 1
    layout["yaxis"+str(tNum)]= dict(
        anchor= "x"+str(i), 
        autorange= True, 
        overlaying= "y"+str(tNum-1),
        domain= [yStart, yStart+yGap], 
        side= "right", 
        type= "linear")
    tNum = tNum + 1
    yStart = yGap+ySpace
    
updatemenus = list([
        #        dict(
        #            x = 0.75,
        #            y = 1.15,
        #            xref = 'paper',
        #            yref = 'paper',
        #            yanchor = 'top',
        #            active = 1,
        #            showactive = False,
        #            buttons = agg_func
        #            ),
                dict(
                    x = 1,
                    y = 1.15,
                    xref = 'paper',
                    yref = 'paper',
                    yanchor = 'top',
                    active = 1,
                    showactive = False,
                    buttons = agg_func2
                    )
                ]
            )

layout['updatemenus'] = updatemenus
fig['layout'] = layout
                              
off.iplot(fig)

########################

Thanks for your help!
Luis
@jonmmease
Copy link
Contributor

Hi @ekn1v ,

This issue you're running into is that the Figure class doesn't support validating and storing transform properties, and this is what make_subplots returns.

One approach you can use is to convert the Figure to a dict before adding transforms using Figure.to_dict(). Then add your transforms to the dict representation of the figure. Then, set the validate argument to False when plotting the figure (e.g. offline.iplot(fig, validate=False)).

As a side note, when working in Python I generally recommend performing transform-like operations in pandas before constructing your traces. This is a bit more flexible and easier to debug (because you can look at the transformed values as pandas dataframes before they get passed to Plotly.js).

@jonmmease
Copy link
Contributor

Closing as duplicate of #587

@ekn1v
Copy link
Author

ekn1v commented Sep 14, 2018 •
edited
Loading

Hi @jonmmease ,

Thank you for helping. I am performing the transform operations in pandas now before constructing the traces.

I am slightly lost on how to convert the Figure to dict and then pass the traces. Can you add some some more details on how to do this?

Thank you so much!

@jonmmease
Copy link
Contributor

If you're no longer trying to set the transform property of the traces, you don't need to worry about converting the figure to a dict (That was just a workaround if you needed to continue using the transform property). If you have more specific questions on your example, feel free to ask a question on the forums at https://community.plot.ly/c/api/python

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@jonmmease @ekn1v