|
 Three Data Sets, one with more values than another |
| Posted by AnthonyB on May-02-2014 00:05 |
|
How should I go about leveling these all out so my graph doesn't look like this:
http://www.tinozplace.com/multiline.png
The outdoor temperature is reported quite often, the indoor temp and AC State only change a every so often. This makes for an uneven amount of data.
What would be the best way to do this? |
| Re: Three Data Sets, one with more values than another |
| Posted by Peter Kwan on May-02-2014 00:25 |
|
Hi AnthonyB,
You may refer to the "Uneven Data Points" sample code as a reference. Basically, when you
add a line layer, you can provide the x-coordinates of the data points using Layer.setXData.
For your case, you can create 3 line layers for the 3 lines, and use different x-coordinates
for the 3 lines. ChartDirector will plot the data points based on the x-coordinates. By
default, it will also automatically determine the x-axis scale (just like it would automatically
determine the y-axis scale), so the Axis.setLabels is not needed.
If you do not provide the x-coordinates, ChartDirector will assume the data points are align
with the axis labels one to one. This would produce the chart you currently have.
Hope this can help.
Regards
Peter Kwan |
| Re: Three Data Sets, one with more values than another |
| Posted by AnthonyB on May-02-2014 01:39 |
|
Hello again...
I think I got all my data sets right but the graph isn't getting what I want..
http://pastebin.com/c0U8u0vt
That's the code, the url for the picture is on the original post for the updated graph. |
| Re: Three Data Sets, one with more values than another |
| Posted by AnthonyB on May-02-2014 02:13 |
|
Ok, I think I see where this is going, sorry for the double post...
data0X = [chartTime(2007, 1, 1), chartTime(2007, 1, 2), chartTime(2007, 1, 5),
chartTime(2007, 1, 7), chartTime(2007, 1, 10), chartTime(2007, 1, 14), chartTime(
2007, 1, 17), chartTime(2007, 1, 18), chartTime(2007, 1, 19), chartTime(2007, 1,
20), chartTime(2007, 1, 21)]
In there it defines the entire amount of time that the graph is going to be valid for.. In my dataset located at www.tinozplace.com/nestlog.txt you can see that has date and time..
When I define this for data0x, what is the best way to automate this? |
| Re: Three Data Sets, one with more values than another |
| Posted by Peter Kwan on May-02-2014 02:51 |
|
Hi AnthonyB,
From your original data, you need to separate them into 6 series:
- data for AC state
- timestamps for AC state
- data for indoor temp
- timetamps for indoor temp
- data for outdoor temp
- timestamps for outdoor temp
Then you can use the 6 series as the data and x-coordinates for the 3 line layers.
Below is what I have just tried based on your original code.
Hope this can help.
Regards
Peter Kwan
import re
import urllib
from pychartdir import *
nestlog = urllib.urlopen('http://www.tinozplace.com/nestlog.txt')
#Get AC State Data
acstate = []
lengths = []
indoortemp = []
outdoortemp = []
for line in nestlog:
if re.match(".+Home_hvac_ac_state.+", line):
match = line.replace('Home_hvac_ac_state', '')
match = match.replace(' ', ' ')
match = match.replace('True', '100')
match = match.replace('False', '0')
match = match.replace('\\n', '')
match = match.replace('\\r', '')
match = match.replace(' ', ', ')
match = match.split(',')
acstate.append(match)
if re.match(".+Home_current_temperature.+", line):
idmatch = line.replace('Home_current_temperature', '')
idmatch = idmatch.replace(' ', ' ')
idmatch = idmatch.replace('\\n', '')
idmatch = idmatch.replace('\\r', '')
idmatch = idmatch.replace(' ', ', ')
idmatch = idmatch.split(',')
indoortemp.append(idmatch)
if re.match(".+weather_current_temp_f.+", line):
odmatch = line.replace('weather_current_temp_f', '')
odmatch = odmatch.replace(' ', ' ')
odmatch = odmatch.replace('\\n', '')
odmatch = odmatch.replace('\\r', '')
odmatch = odmatch.replace(' ', ', ')
odmatch = odmatch.split(',')
outdoortemp.append(odmatch)
#print odmatch
def parseDTV(date_time_value) :
xCoor = []
data = []
for myDate, myTime, myValue in date_time_value:
yyyy, mm, dd = myDate.split('-')
hh, nn, ss = myTime.split(':')
xCoor.append(chartTime(yyyy, mm, dd, hh, nn, ss))
data.append(myValue)
return xCoor, data
ACXCoor, ACData = parseDTV(acstate)
IDXCoor, IDData = parseDTV(indoortemp)
ODXCoor, ODData = parseDTV(outdoortemp)
# Create a XYChart object of size 600 x 400 pixels. Use a vertical gradient color
# from light blue (99ccff) to white (ffffff) spanning the top 100 pixels as
# background. Set border to grey (888888). Use rounded corners. Enable soft drop
# shadow.
c = XYChart(600, 400)
c.setBackground(c.linearGradientColor(0, 0, 0, 100, 0x99ccff, 0xffffff), 0x888888)
c.setRoundedFrame()
c.setDropShadow()
# Add a title using 18 pts Times New Roman Bold Italic font. Set top margin to 16
# pixels.
c.addTitle("Nest Daata", "timesbi.ttf", 18).setMargin2(0, 0, 16, 0)
# Set the plotarea at (60, 80) and of 510 x 275 pixels in size. Use transparent
# border and dark grey (444444) dotted grid lines
plotArea = c.setPlotArea(60, 80, 510, 275, -1, -1, Transparent, c.dashLineColor(
0x444444, 0x0101), -1)
# Add a legend box where the top-center is anchored to the horizontal center of the
# plot area at y = 45. Use horizontal layout and 10 points Arial Bold font, and
# transparent background and border.
legendBox = c.addLegend(plotArea.getLeftX() + plotArea.getWidth() / 2, 45, 0,
"arialbd.ttf", 10)
legendBox.setAlignment(TopCenter)
legendBox.setBackground(Transparent, Transparent)
# Set x-axis tick density to 75 pixels and y-axis tick density to 30 pixels.
# ChartDirector auto-scaling will use this as the guidelines when putting ticks on
# the x-axis and y-axis.
c.yAxis().setTickDensity(30)
c.xAxis().setTickDensity(75)
# Set all axes to transparent
c.xAxis().setColors(Transparent)
c.yAxis().setColors(Transparent)
# Set the x-axis margins to 15 pixels, so that the horizontal grid lines can extend
# beyond the leftmost and rightmost vertical grid lines
c.xAxis().setMargin(15, 15)
# Set axis label style to 8pts Arial Bold
c.xAxis().setLabelStyle("arialbd.ttf", 8)
c.yAxis().setLabelStyle("arialbd.ttf", 8)
c.yAxis2().setLabelStyle("arialbd.ttf", 8)
# Add axis title using 10pts Arial Bold Italic font
c.yAxis().setTitle("Backlog in USD millions", "arialbi.ttf", 10)
# Add the first data series
layer0 = c.addLineLayer2()
layer0.addDataSet(ACData, 0xff0000, "AC State").setDataSymbol(GlassSphere2Shape,
11)
layer0.setXData(ACXCoor)
layer0.setLineWidth(3)
# Add the second data series
layer1 = c.addLineLayer2()
layer1.addDataSet(IDData, 0x00ff00, "Indoor Temp").setDataSymbol(GlassSphere2Shape,
11)
layer1.setXData(IDXCoor)
layer1.setLineWidth(3)
# Add the third data series
layer2 = c.addLineLayer2()
layer2.addDataSet(ODData, 0xff6600, "Outdoor
Temp").setDataSymbol(GlassSphere2Shape, 11)
layer2.setXData(ODXCoor)
layer2.setLineWidth(3)
# Output the chart
c.makeChart("C:\\\\multiline.png")
print 'Chart Saved' |
| Re: Three Data Sets, one with more values than another |
| Posted by AnthonyB on May-02-2014 08:00 |
|
While I was waiting for a reply, I did some more with this graph, As you'll see though, the Outdoor temperature stays the same..
I have investigated it further and it seems that I don't have any "middle" values for the one with the most updates (outdoor temp) so, it only gives the first two values.
http://pastebin.com/DsVVF5xW
I tried your function out and it worked well for a single value, but not all of the values in the variable.
I am a little lost on how to continue. |
| Re: Three Data Sets, one with more values than another |
| Posted by AnthonyB on May-02-2014 09:47 |
|
Thanks to your function, I was able to figure it out on my own.
Thanks so much for your help. |
|
Forum Home
Search
Message List
Post Message