Re: Week and Month plots not being generated?

[Henry Denston]

On Thursday, September 12, 2019 at 12:41:55 AM UTC+2, Henry Denston wrote:

Hi,

I'm playing around with a customized weewx version (adapted to a custom weather-station) and ran the framework for the first time on the final system for testing.

Everything working great except I noticed that the weekly and montly plots do not include data.

I'm using a modified version of the Seasons skin.

The daily and yearly plots seem to work great and no weewx error or warning in the log.

Only the week and month plot stay empty (only the timestamp of plot generation at the bottom of the respective plot is updated).

Can you think of a reason why this might happen?

The system runs for two days already so I think the week plot should contain some data by now (even the year plot includes some already)?

In the skin.conf file this is the configuration for one observation type:

    [[week_images]]
        x_label_format = %d
        bottom_label_format = %x %X
        time_length = 604800 # 7 days
        aggregate_type = avg
        aggregate_interval = 7200 # generate plot every 2 hours for week plots  
 show_daynight = true
 
        [[[weektempdew]]]
            [[[[outTemp]]]]

Would appreciate any hint, where to start to look for this issue.

Is it possible I still have to wait longer for these plots to be filled with data?

Thank you very much in advance, henry.

[gjr80]

Hi,

Are you sure the plots are being generated without data? Under some circumstances plots can be generated with data points only (ie no connecting lines between the data points) and to a casual observer they can look like there is not data being plotted at all (especially if there are only a few data points). Can you post the plots and the entire [ImageGenerator] stanza from skin.conf. Is your site publicly accessible via the internet, if so a link would help.

Gary

[Henry Denston]

Hi Gary,

thank you very much for your reply.

The website is not public at the moment but maybe I can make it public soon.

Below you can find the code for the entire [ImageGenerator] stanza.

How can I make sure that the plots are not empty but just the data points are not connected?

Can I check smth. in the log to make sure?

(I can not post any plots right now as I do not have access to it atm but will as soon as possible).

# The ImageGenerator creates image plots of data.


[ImageGenerator]


    # This section lists all the images to be generated, what SQL types are to
    # be included in them, along with many plotting options. There is a default
    # for almost everything. Nevertheless, values for most options are included
    # to make it easy to see and understand the options.
    #
    # Fonts can be anything accepted by the Python Imaging Library (PIL), which
    # includes truetype (.ttf), or PIL's own font format (.pil). See
    # http://www.pythonware.com/library/pil/handbook/imagefont.htm for more
    # details.  Note that "font size" is only used with truetype (.ttf)
    # fonts. For others, font size is determined by the bit-mapped size,
    # usually encoded in the file name (e.g., courB010.pil). A relative path
    # for a font is relative to the SKIN_ROOT.  If a font cannot be found,
    # then a default font will be used.
    #
    # Colors can be specified any of three ways:
    #   1. Notation 0xBBGGRR;
    #   2. Notation #RRGGBB; or
    #   3. Using an English name, such as 'yellow', or 'blue'.
    # So, 0xff0000, #0000ff, or 'blue' would all specify a pure blue color.
   
    image_width = 800
    image_height = 300
    image_background_color = 0x262628
   
    chart_background_color = 0x161618
    chart_gridline_color = 0x3a3a3b


    # Setting to 2 or more might give a sharper image with fewer jagged edges
    anti_alias = 4


    top_label_font_path = font/newfont.ttf
    top_label_font_size = 20

    unit_label_font_path = font/newfontr.ttf
    unit_label_font_size = 20
    unit_label_font_color = 0x79C8EB


    bottom_label_font_path = font/newfont.ttf

    bottom_label_font_size = 12
    bottom_label_font_color = 0x2B6DA2
    bottom_label_offset = 20


    axis_label_font_path = font/newfont.ttf
    axis_label_font_size = 12
    axis_label_font_color = 0x2977D3
   
    # Options for the compass rose, used for progressive vector plots
    rose_label = N
    rose_label_font_path = font/newfont.ttf
    rose_label_font_size  = 15
    rose_label_font_color = 0x2977D3


    # Default colors for the plot lines. These can be overridden for
    # individual lines using option 'color'.
    chart_line_colors = 0x79C8EB, 0xFFFFFF, 0x42b444, 0xb4b444, 0xb442b6
   
    # Default fill colors for bar charts. These can be overridden for
    # individual bar plots using option 'fill_color'.
    chart_fill_colors = 0x79C8EB, 0xFFFFFF, 0x72c474, 0xc4c474, 0xc472c6  
   
    # Type of line. Options are 'solid' or 'none'.
    line_type = 'solid'


    # Size of marker in pixels
    marker_size = 8


    # Type of marker. Options are 'cross', 'x', 'circle', 'box', or 'none'.
    marker_type ='none'
       
    # The following option merits an explanation. The y-axis scale used for
    # plotting can be controlled using option 'yscale'. It is a 3-way tuple,
    # with values (ylow, yhigh, min_interval). If set to "None", a parameter is
    # set automatically, otherwise the value is used. However, in the case of
    # min_interval, what is set is the *minimum* y-axis tick interval.
    yscale = None, None, None


    # For progressive vector plots, you can choose to rotate the vectors.
    # Positive is clockwise.
    # For my area, westerlies overwhelmingly predominate, so by rotating
    # positive 90 degrees, the average vector will point straight up.
    vector_rotate = 90


    # This defines what fraction of the difference between maximum and minimum
    # horizontal chart bounds is considered a gap in the samples and should not
    # be plotted.
    line_gap_fraction = 0.01


    # This controls whether day/night bands will be shown. They only look good
    # on plots wide enough to show individual days such as day and week plots.
    show_daynight = true
    # These control the appearance of the bands if they are shown.
    # Here's a monochrome scheme:
    daynight_day_color   = 0x262628
    daynight_night_color = 0x161618
    daynight_edge_color  = 0x212123


    # What follows is a list of subsections, each specifying a time span, such
    # as a day, week, month, or year. There's nothing special about them or
    # their names: it's just a convenient way to group plots with a time span
    # in common. You could add a time span [[biweek_images]] and add the
    # appropriate time length, aggregation strategy, etc., without changing
    # any code.
    #
    # Within each time span, each sub-subsection is the name of a plot to be
    # generated for that time span. The generated plot will be stored using
    # that name, in whatever directory was specified by option 'HTML_ROOT'
    # in weewx.conf.
    #
    # With one final nesting (four brackets!) is the sql type of each line to
    # be included within that plot.
    #
    # Unless overridden, leaf nodes inherit options from their parent
   
    # Default plot parameters
    plot_type = line
    aggregate_type = none
    width = 1
    time_length = 86400 # 24 hours
 x_label_spacing = 1


    [[day_images]]
        x_label_format = %H:%M
        bottom_label_format = %x %X
        time_length = 97200 # 27 hours
        aggregate_interval = 300 # every 5 minutes for day plots  
 show_daynight = true
 
        [[[daytempdew]]]
            [[[[outTemp]]]]
       
        [[[daytempfeel]]]
            [[[[windchill]]]]
       
        [[[daywind]]]
            [[[[windSpeed]]]]
            [[[[windGust]]]]


        [[[dayradiation]]]
            [[[[radiation]]]]


        [[[dayradiationD]]]
            [[[[radiationD]]]]

    [[week_images]]
        x_label_format = %d
        bottom_label_format = %x %X
        time_length = 604800 # 7 days
        aggregate_type =

avg
        aggregate_interval = 7200 # every 2 hours for week plots  

 show_daynight = true
 
        [[[weektempdew]]]
            [[[[outTemp]]]]

       
        [[[weektempfeel]]]
            [[[[windchill]]]]
       
        [[[weekwind]]]
            [[[[windSpeed]]]]
            [[[[windGust]]]]
                aggregate_type = max


        [[[weekradiation]]]
            [[[[radiation]]]]


        [[[weekradiationD]]]
            [[[[radiationD]]]]


    [[month_images]]

        x_label_format = %d
        bottom_label_format = %x %

X
        time_length = 2592000 # 30 days
        aggregate_type = avg
        aggregate_interval = 28800 # every 8 hours for month plots  
        show_daynight = false
       
        [[[monthtempdew]]]
            [[[[outTemp]]]]
       
        [[[monthtempfeel]]]
            [[[[windchill]]]]
       
        [[[monthwind]]]
            [[[[windSpeed]]]]
            [[[[windGust]]]]
                aggregate_type = max


        [[[monthradiation]]]
            [[[[radiation]]]]


        [[[monthradiationD]]]
            [[[[radiationD]]]]


    [[year_images]]
        x_label_format = %m/%d
        bottom_label_format = %x %X
        time_length = 31536000 # 365 days
        aggregate_type = avg
        aggregate_interval = 86400 # once a day for year plots  
        show_daynight = false


        [[[yeartempdew]]]
            [[[[outTemp]]]]
       
        [[[yeartempfeel]]]
            [[[[windchill]]]]
               
        [[[yearwind]]]
            [[[[windSpeed]]]]
            [[[[windGust]]]]
                aggregate_type = max


        [[[yearradiation]]]
            [[[[radiation]]]]


        [[[yearradiationD]]]
            [[[[radiationD]]]]


        # Plot of high/low temperatures
        [[[yearhilow]]]
            [[[[hi]]]]
                data_type = outTemp
                aggregate_type = max
                label = High
            [[[[low]]]]
                data_type = outTemp
                aggregate_type = min
                label = Low Temperature

[gjr80]

Hi Henry,

The [ImageGenerator] stanza looks fine. There is no practical way to confirm if you have dots in your plots other than careful observation of the generated plot. Have a read of the section ‘Dots in your plots’ (http://weewx.com/docs/usersguide.htm#dots_in_plots) in the Users Guide, it’s possible you may need to adjust the value of the line_gap_fraction option. Of course this may not be the problem at all, we don’t know yet whether you do/do not have dots in your plots.

Gary