Streaming circle shapes graph #268

ghost · 2015-12-31T08:05:30Z

Hi - is there a streaming example for a circle shape plot? In this example : https://plot.ly/python/shapes/#circle

trace0 = go.Scatter(
x=[1.5, 3.5],
y=[0.75, 2.5],
text=['Unfilled
Circle', 'Filled
Circle'],
mode='text',
)

How do I declare this for streaming? Will x and y be 2 dimensional arrays? For a line graph I normally declare like this:
trace_ctr = Scatter(
x=[],
y=[]
)
How do I do it for a circle?

theengineear · 2015-12-31T20:44:03Z

@cloaked123, here's a full, working example of a stream with shapes (circles in this case):

import random
import time

import colorlover as cl

import plotly.plotly as py
import plotly.graph_objs as go

# Define a couple functions to make the important bits more readable
def get_shape(color, type='circle'):
    x_center = random.random() * 10
    y_center = random.random() * 10
    dx = random.random() * 2
    dy = random.random() * 2
    x0 = x_center - dx/2.0
    y0 = y_center - dy/2.0
    x1 = x_center + dx/2.0
    y1 = y_center + dy/2.0
    return {'type': type, 'xref': 'x', 'yref': 'y', 'fillcolor': color,
            'x0': x0, 'x1': x1, 'y0': y0, 'y1': y1}

def move_shape(shape):
    dx = shape['x1'] - shape['x0']
    dy = shape['y1'] - shape['y0']
    x_center = (shape['x1'] + shape['x0']) / 2
    y_center = (shape['y1'] + shape['y0']) / 2
    x_center += (random.random() - .5) / 4
    y_center += (random.random() - .5) / 4
    dx += (random.random() - .5) / 4
    dy += (random.random() - .5) / 4
    shape['x0'] = x_center - dx/2
    shape['y0'] = y_center - dy/2
    shape['x1'] = x_center + dx/2
    shape['y1'] = y_center + dy/2

# Make the initial plot.
token = py.get_credentials()['stream_ids'][0]
num_shapes = 50
mod = 10
colors = cl.scales[str(mod)]['div']['RdYlBu']

# Traces hold stream information. In this case, we *at least*
# need a place holder trace to hold the streaming target.
trace0 = go.Scatter(stream={'token': token})

# The layout holds the shapes, so we dump all the shape information there.
# We'll update the layout via streaming later.
layout = {'width': 800, 'height': 800, 'autosize': False,
          'xaxis': {'range': [-5, 15]}, 'yaxis': {'range': [-5, 15]},
          'shapes': [get_shape(colors[i % mod]) for i in range(num_shapes)]}

fig = {'data': [trace0], 'layout': layout}
print py.plot(fig, filename='messin', auto_open=False)

# This is the important part. You can use the `layout` kwarg in the call
# to Stream.write(). Note that you *need* to send along an empty trace or
# the Python client will think you made a mistake.
s = py.Stream(token)
s.open()
for _ in range(500):
    for shape in layout['shapes']:
        move_shape(shape)
    s.write({}, layout=layout)
    time.sleep(.1)
s.close()

Helpful? The gist is that you need to use the layout kwarg in the Stream.write() function.

ghost · 2016-01-01T02:29:11Z

Wow this is awesome. Thanks a lot!

Sent from my iPhone

On Thu, Dec 31, 2015 at 12:44 PM -0800, "Andrew" [email protected] wrote:

@cloaked123, here's a full, working example of a stream with shapes (circles in this case):

import random
import time

import colorlover as cl

import plotly.plotly as py
import plotly.graph_objs as go

Define a couple functions to make the important bits more readable

def get_shape(color, type='circle'):
x_center = random.random() * 10
y_center = random.random() * 10
dx = random.random() * 2
dy = random.random() * 2
x0 = x_center - dx/2.0
y0 = y_center - dy/2.0
x1 = x_center + dx/2.0
y1 = y_center + dy/2.0
return {'type': type, 'xref': 'x', 'yref': 'y', 'fillcolor': color,
'x0': x0, 'x1': x1, 'y0': y0, 'y1': y1}

def move_shape(shape):
dx = shape['x1'] - shape['x0']
dy = shape['y1'] - shape['y0']
x_center = (shape['x1'] + shape['x0']) / 2
y_center = (shape['y1'] + shape['y0']) / 2
x_center += (random.random() - .5) / 4
y_center += (random.random() - .5) / 4
dx += (random.random() - .5) / 4
dy += (random.random() - .5) / 4
shape['x0'] = x_center - dx/2
shape['y0'] = y_center - dy/2
shape['x1'] = x_center + dx/2
shape['y1'] = y_center + dy/2

Make the initial plot.

token = py.get_credentials()['stream_ids'][0]
num_shapes = 50
mod = 10
colors = cl.scales[str(mod)]['div']['RdYlBu']

Traces hold stream information. In this case, we at least

need a place holder trace to hold the streaming target.

trace0 = go.Scatter(stream={'token': token})

The layout holds the shapes, so we dump all the shape information there.

We'll update the layout via streaming later.

layout = {'width': 800, 'height': 800, 'autosize': False,
'xaxis': {'range': [-5, 15]}, 'yaxis': {'range': [-5, 15]},
'shapes': [get_shape(colors[i % mod]) for i in range(num_shapes)]}

fig = {'data': [trace0], 'layout': layout}
print py.plot(fig, filename='messin', auto_open=False)

This is the important part. You can use the `layout` kwarg in the call

to Stream.write(). Note that you need to send along an empty trace or

the Python client will think you made a mistake.

s = py.Stream(token)
s.open()
for _ in range(500):
for shape in layout['shapes']:
move_shape(shape)
s.write({}, layout=layout)
time.sleep(.1)
s.close()

Helpful? The gist is that you need to use the layout kwarg in the Stream.write() function.

—
Reply to this email directly or view it on GitHub.

ghost · 2016-01-06T18:31:39Z

Hello -
Could you please help me with another issue:

In the following plot - how do I make the star marker appear above the
circle shapes instead of behind?

https://plot.ly/23/~cloaked/

Thanks
-Ravi

On Thu, Dec 31, 2015 at 3:44 PM, Andrew [email protected] wrote:

@cloaked123 https://github.com/cloaked123, here's a full, working
example of a stream with shapes (circles in this case):

import random
import time

import colorlover as cl

import plotly.plotly as py
import plotly.graph_objs as go

Define a couple functions to make the important bits more readable

def get_shape(color, type='circle'):
x_center = random.random() * 10
y_center = random.random() * 10
dx = random.random() * 2
dy = random.random() * 2
x0 = x_center - dx/2.0
y0 = y_center - dy/2.0
x1 = x_center + dx/2.0
y1 = y_center + dy/2.0
return {'type': type, 'xref': 'x', 'yref': 'y', 'fillcolor': color,
'x0': x0, 'x1': x1, 'y0': y0, 'y1': y1}

def move_shape(shape):
dx = shape['x1'] - shape['x0']
dy = shape['y1'] - shape['y0']
x_center = (shape['x1'] + shape['x0']) / 2
y_center = (shape['y1'] + shape['y0']) / 2
x_center += (random.random() - .5) / 4
y_center += (random.random() - .5) / 4
dx += (random.random() - .5) / 4
dy += (random.random() - .5) / 4
shape['x0'] = x_center - dx/2
shape['y0'] = y_center - dy/2
shape['x1'] = x_center + dx/2
shape['y1'] = y_center + dy/2

Make the initial plot.

token = py.get_credentials()['stream_ids'][0]
num_shapes = 50
mod = 10
colors = cl.scales[str(mod)]['div']['RdYlBu']

Traces hold stream information. In this case, we at least

need a place holder trace to hold the streaming target.

trace0 = go.Scatter(stream={'token': token})

The layout holds the shapes, so we dump all the shape information there.

We'll update the layout via streaming later.

layout = {'width': 800, 'height': 800, 'autosize': False,
'xaxis': {'range': [-5, 15]}, 'yaxis': {'range': [-5, 15]},
'shapes': [get_shape(colors[i % mod]) for i in range(num_shapes)]}

fig = {'data': [trace0], 'layout': layout}
print py.plot(fig, filename='messin', auto_open=False)

This is the important part. You can use the layout kwarg in the call

to Stream.write(). Note that you need to send along an empty trace or

the Python client will think you made a mistake.

s = py.Stream(token)
s.open()
for _ in range(500):
for shape in layout['shapes']:
move_shape(shape)
s.write({}, layout=layout)
time.sleep(.1)
s.close()

Helpful? The gist is that you need to use the layout kwarg in the
Stream.write() function.

—
Reply to this email directly or view it on GitHub
#268 (comment)
.

theengineear · 2016-01-06T19:00:54Z

@cloaked123 , there's no current way to do that right now. There's an open issue for it in the plotly.js repo though:

plotly/plotly.js#148

theengineear · 2016-01-06T19:02:04Z

Closing this down as it's not related to the python api at this point.

ghost · 2016-01-06T19:21:37Z

Thanks

On Wed, Jan 6, 2016 at 2:00 PM, Andrew [email protected] wrote:

@cloaked123 https://github.com/cloaked123 , there's no current way to
do that right now. There's an open issue for it in the plotly.js repo
though:

plotly/plotly.js#148 plotly/plotly.js#148

—
Reply to this email directly or view it on GitHub
#268 (comment)
.

theengineear closed this as completed Jan 6, 2016

cldougl mentioned this issue Feb 25, 2016

improve python streaming docs #183

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Streaming circle shapes graph #268

Streaming circle shapes graph #268

ghost commented Dec 31, 2015

theengineear commented Dec 31, 2015

Uh oh!

ghost commented Jan 1, 2016

Uh oh!

ghost commented Jan 6, 2016

Define a couple functions to make the important bits more readable

Make the initial plot.

Traces hold stream information. In this case, we at least

need a place holder trace to hold the streaming target.

The layout holds the shapes, so we dump all the shape information there.

We'll update the layout via streaming later.

This is the important part. You can use the `layout` kwarg in the call

to Stream.write(). Note that you need to send along an empty trace or

the Python client will think you made a mistake.

Uh oh!

theengineear commented Jan 6, 2016

Uh oh!

theengineear commented Jan 6, 2016

Uh oh!

ghost commented Jan 6, 2016

Uh oh!

Uh oh!

Streaming circle shapes graph #268

Streaming circle shapes graph #268

Comments

ghost commented Dec 31, 2015

theengineear commented Dec 31, 2015

Uh oh!

ghost commented Jan 1, 2016

Define a couple functions to make the important bits more readable

Make the initial plot.

Traces hold stream information. In this case, we at least

need a place holder trace to hold the streaming target.

The layout holds the shapes, so we dump all the shape information there.

We'll update the layout via streaming later.

This is the important part. You can use the layout kwarg in the call

to Stream.write(). Note that you need to send along an empty trace or

the Python client will think you made a mistake.

Uh oh!

ghost commented Jan 6, 2016

Define a couple functions to make the important bits more readable

Make the initial plot.

Traces hold stream information. In this case, we at least

need a place holder trace to hold the streaming target.

The layout holds the shapes, so we dump all the shape information there.

We'll update the layout via streaming later.

This is the important part. You can use the layout kwarg in the call

to Stream.write(). Note that you need to send along an empty trace or

the Python client will think you made a mistake.

Uh oh!

theengineear commented Jan 6, 2016

Uh oh!

theengineear commented Jan 6, 2016

Uh oh!

ghost commented Jan 6, 2016

Uh oh!

This is the important part. You can use the `layout` kwarg in the call

This is the important part. You can use the `layout` kwarg in the call