Python Bar Chart

[RB]

Hi all,

I'm experimenting with python chart (plotly is available here so more interactive charts possible then with R charts).

The R chart i want to re-create in python:

#Get activity (used to get the date)
activity <- GC.activity.metrics()
#Get zone information at this actvity date
zone <- GC.athlete.zones(date=activity$date, sport="bike")

#Get intervals USER if that one is not find get all intervals
intervals <- GC.activity.intervals(type="USER")
type_title <- "USER"
if (is.data.frame(intervals) && nrow(intervals)==0){
  type_title <- "ALL"
  intervals <- GC.activity.intervals()
}

#Identify for every interval the zone color
breaks <- c(unlist(zone$zoneslow))
interval_colors <- c(unlist(zone$zonescolor))[findInterval(intervals$Average_Power, vec=breaks)]

#Define chart title
title <- paste0("Average Power per Interval (CP:",zone$cp,") Interval Type=", type_title)

#Define margin (when interval names are not printed propably the names are to long change bottom value(first parameter))
par(mar=c(10,5,5,0))


barplot <- barplot(intervals$Average_Power, axes = FALSE, axisnames = FALSE, col=interval_colors, width=intervals$Duration,
main=title,
ylim=c(0,max(intervals$Average_Power)*1.1)
)

#Add horizontal line at 0
abline(h=0)

#Add interval names at X-as
text(barplot,
     mean(intervals$Average_Power)*-0.05,
     labels = intervals$name,
     srt = 90,
     xpd = TRUE,
     adj=0,
)

#Add percentage labels
avg_power_pct <- round((intervals$Average_Power / zone$cp)*100,1)
avg_power_pct_label<- paste(avg_power_pct, "%")
text(barplot,
     intervals$Average_Power+10,
     labels = avg_power_pct_label,
)
axis(2)


#Add legend
zone_names = c("Z1", "Z2", "Z3", "Z4", "Z5", "Z6", "Z7")
legend <- paste(zone_names, "(" , breaks, ")")
legend("topleft",
  inset=0.02,
  legend=legend,
  fill=c(unlist(zone$zonescolor)),
)

R Result:

Python Chart:

import bisect
import plotly
import plotly.graph_objs as go
import numpy as np


#Get activity (used to get the date)
activity = GC.activityMetrics()
#Get zone information at this actvity date
zone = GC.athleteZones(date=activity["date"], sport="bike")

#Get intervals USER if that one is not find get all intervals
intervals = GC.activityIntervals(type="USER")
type_title = "USER"
if not intervals["name"]:
  type_title = "ALL"
  intervals = GC.activityIntervals()
 

#Identify for every interval the zone color
breaks = zone["zoneslow"][0]
zone_colors = zone["zonescolor"][0]
interval_color_list = []
avg_power_pct =[]
for interval in intervals["Average_Power"]:
  id = bisect.bisect_left(breaks, interval)
  interval_color_list.append(zone_colors[id-1])
  avg_power_pct.append(str(round((interval / zone["cp"][0])*100,1))+"%")

#Define chart title
title = "Average Power per Interval (CP:" + str(zone["cp"]) + ") Interval Type=" + str(type_title)

#Add percentage labels
zone_names = ["Z1", "Z2", "Z3", "Z4", "Z5", "Z6", "Z7"]
legend = []
zone_index=1
for zone in breaks:
  legend.append("Z" + str(zone_index) + "(" + str(zone) + ")")
  zone_index += 1

xx = np.asarray(intervals["name"])
yy = np.asarray(intervals["Average_Power"])
ff = np.arange(1,len(xx))

#print(ff)
#print(xx)
#print(yy)
data = []
print(duration)
data.append(go.Bar(
            x=xx,
            y=yy,
            name="avg power",
#            width=intervals["Duration"],
            text=avg_power_pct,
            textposition = 'auto',
            showlegend=False,
            marker=dict(
              color=interval_color_list),
        ))

#workaround to add legend.
for i in range(len(breaks)):
  data.append(
      go.Scatter(x=[None], y=[None],
      mode='markers',
      marker=dict(size=10,  color=zone_colors[i]),
      legendgroup=str(legend[i]),
      showlegend=True,
      name=zone_names[i])
)
layout = go.Layout(
title=go.layout.Title(
        text=title,
        xref='paper',
        x=0
    ),
)

fig = go.Figure(data=data, layout=layout)
barplot = plotly.offline.plot(fig, auto_open = False, filename="D:/temp.html")

## Load Plot
GC.webpage(barplot)

Python result in :

So Questions(Help needed on):

1. Why are the Rest interval stacked (it is not a stacked bar)?

2. How to set the bar width (bar width equal to duation of the interval)?

3. I need to set the filename on a absolote path else it doesn't show (while it is creating the html my filesystem), why or is this a bug?

4. What is the roadmap of golden cheetah more R oriented or move to python?

Thanks in advance kind regards,

RB

[Marco1967]

super thank you!

yy = np.asarray(intervals<span style="color: #660;" class="styled-by-pre

[RB]

Investigated some more so some answers on my own question

So Questions(Help needed on):

1. Why are the Rest interval stacked (it is not a stacked bar)?

    x values with the same name are mapped on each other on the x-as

2. How to set the bar width (bar width equal to duation of the interval)?

    Not possible with plotly (with bar charts), i managed to do with shapes

I would still like answer on:

3. I need to set the filename on a absolote path else it doesn't show (while it is creating the html on my filesystem), why or is this a bug?

4. What is the roadmap of golden cheetah more R oriented or move to python?

This is what I ended up with plotly and shapes:

Chart code:

import bisect
import plotly
import plotly.graph_objs as go
import numpy as np


#Get activity (used to get the date)
activity = GC.activityMetrics()
#Get zone information at this actvity date
zone = GC.athleteZones(date=activity["date"], sport="bike")

#Get intervals USER if that one is not find get all intervals

intervals = GC.activityIntervals(type="USER")
type_title = "USER"
if not intervals["name"]:
  type_title = "ALL"
  intervals = GC.activityIntervals()
 

#Identify for every interval the zone color
breaks = zone["zoneslow"][0]
zone_colors = zone["zonescolor"][0]

interval_colors = []

avg_power_pct =[]
for interval in intervals["Average_Power"]:
  id = bisect.bisect_left(breaks, interval)

  interval_colors.append(zone_colors[id-1])

  avg_power_pct.append(str(round((interval / zone["cp"][0])*100,1))+"%")


#Define chart title
title = "Average Power per Interval (CP:" + str(zone["cp"]) + ") Interval Type=" + str(type_title)

#Add percentage labels
zone_names = ["Z1", "Z2", "Z3", "Z4", "Z5", "Z6", "Z7"]
legend = []
zone_index=1
for zone in breaks:
  legend.append("Z" + str(zone_index) + "(" + str(zone) + ")")
  zone_index += 1

lap_names = np.asarray(intervals["name"])
watts_y = np.asarray(intervals["Average_Power"])
x = np.asarray(intervals["start"])
duration = np.asarray(intervals["Duration"])

trace0 = go.Scatter(
    x=x,
    y=watts_y,
    mode='text',
    showlegend=False,

)

data = [trace0]
for i in np.arange(0,len(legend)):

    data.append(go.Scatter(
        x=[None],
        y=[None],
        mode='markers',
        marker=dict(size=10, color=zone_colors[i]),

        legendgroup=legend[i],
        showlegend=True,
        name=legend[i])
    )


# Create rectangles per interval
shapes = []
annotations = []
for i in np.arange(0,len(lap_names)):
    shapes.append(
       {
        'type': 'rect',
        'x0': x[i],
        'y0': 0,
        'x1': x[i]+duration[i],
        'y1': watts_y[i],
        'fillcolor': interval_colors[i],
    })
    annotations.append(
        dict(
            x=x[i] + (duration[i]/2),
            y=watts_y[i],
            xref='x',
            yref='y',
            text=str(lap_names[i]) + "<br>" + str(avg_power_pct[i]) + "<br>" + str(duration[i]) + "s",
            showarrow=True,
            arrowhead=7,
            ax=0,
            ay=-40
        ))


layout = go.Layout(
    title = title,
    xaxis = dict(
        range=[0, max(x)],
        showgrid=True,
        autorange=True,
        zeroline=False,
        showline=False,
        ticks='',
        showticklabels=False

    ),
    yaxis = dict(
        range=[0,  max(watts_y)]
    ),
    margin=go.layout.Margin(
        l=50,
        r=50,
        b=100,
        t=150,
        pad=4
    ),
    shapes=shapes,
    annotations=annotations,

)

fig = go.Figure(data=data, layout=layout)

plot = plotly.offline.plot(fig, auto_open=False, filename='D:/temp_chart.html')
## Load Plot
GC.webpage(plot)

[Ale Martinez - No direct email please]

El viernes, 28 de junio de 2019, 12:37:20 (UTC-6), RB escribió:
> Investigated some more so some answers on my own question
>
>
>
>
> So Questions(Help needed on):
> 1. Why are the Rest interval stacked (it is not a stacked bar)?
>     x values with the same name are mapped on each other on the x-as
>
> 2. How to set the bar width (bar width equal to duation of the interval)?
>     Not possible with plotly (with bar charts), i managed to do with shapes
>
>
> I would still like answer on:
>
> 3.
> I need to set the filename on a absolote path else it doesn't show
> (while it is creating the html on my filesystem), why or is this a bug?4.

The html file creation is necessary to use plotly offline, if you want to make your code portable the tempfile Python module may help

> What is the roadmap of golden cheetah more R oriented or move to python?

The objective was to have a similar API and functionality as far as possible but there are limitations with multi threading on R so Scripting in formulas (useful for complex custom metrics) is limited to Python.

> This is what I ended up with plotly and shapes:
>
>
>

> Chart code:
>
>
> import bisect
> import plotly
> import plotly.graph_objs as go
> import numpy as np
>
>
> #Get activity (used to get the date)
> activity = GC.activityMetrics()
> #Get zone information at this actvity date
> zone = GC.athleteZones(date=activity["date"], sport="bike")
>
> #Get intervals USER if that one is not find get all intervals
>
> intervals = GC.activityIntervals(type="USER")
> type_title = "USER"
> if not intervals["name"]:
>   type_title = "ALL"
>   intervals = GC.activityIntervals()
>  
>
> #Identify for every interval the zone color
> breaks = zone["zoneslow"][0]
> zone_colors = zone["zonescolor"][0]
> interval_colors = []
> avg_power_pct =[]
> for interval in intervals["Average_Power"]:
>   id = bisect.bisect_left(breaks, interval)
>   interval_colors.append(zone_colors[id-1])

>   avg_power_pct.append(str(round((interval / zone["cp"][0])*100,1))+"%")m

Nice example, if you want to contribute the wiki is open.

[RB]

Nice example, if you want to contribute the wiki is open.

Thx, no problem to add the example to the wiki. Is there a example place on the wiki or should i place it in page: https://github.com/GoldenCheetah/GoldenCheetah/wiki/UG_Special-Topics_Working-with-Python

Also uploaded a refined chart to the cloud.

[Mark Miller]

Now that I got python working I tried these examples here and I can get the R Chart to show in GC but the the Python code does not display a chart in GC. It generates an html file on my system but it will not show in GC. I'm using macOS and 3.5-RC1

[mlt.m...@googlemail.com]

Its working for my on v3.6-DEV2006 if change the last two lines.

From

plot = plotly.offline.plot(fig, auto_open=False, filename='D:/temp_chart.html')
## Load Plot
GC.webpage(plot)

To

plot = plotly.offline.plot(fig, auto_open=False, filename=temp_file.name)

## Load Plot

GC.webpage(pathlib.Path(temp_file.name).as_uri())

[kub3r4]

If running in Windows and have disk "D:" then it would work. But you would need to open "temp_chart.html" in browser.

But if you change only "auto_open" from False to True, it will automatically open "temp_chart.html" in your browser.

[P Stroud-Baranda]

Thank you for the code and explanations.