Differences

This shows you the differences between two versions of the page.

--- historical_graphs [2018/12/24 16:17] – emozolyak
+++ historical_graphs [2023/03/23 09:56] (current) – [Log data size considerations] emozolyak
@@ Line 1: / Line 1: @@
+{{ network:menu-icon-graphs.png?nolink&40|}}
 ====== Historical graphs ======
@@ Line 10: / Line 11: @@
 This one is shown on a small time scale:
-{{::graph_example_pure_curve_steps.png?direct}}
+{{network:graph_example_pure_curve_steps.png?direct}}
 This one is shown on a large time scale:
-{{::graph_example_aerial_view_steps.png?direct}}
+{{network:graph_example_aerial_view_steps.png?direct}}
 Elements of the graph:
   *(1) - calendar for selecting date/period of view
-  *(2) - Graph title
+  *(2) - graph title
-  *(3) - title of a register ( or its legend in the graph settings) and respective values in the selected point on the curve. Because with large time scale there can be more log data in a given pixel of the curve (which corresponds to one X-axis division formed as time span / grapsh area in pixels )  3 values are shown - mininal, average and maximal. When there are values in the log database for all points of curve in a given time scale, all three values will be the same ( and the curve will look "pure").
+  *(3) - title of a register ( or its legend in the graph settings) and respective values in the selected point on the curve. Because with large time scale there can be more log data in a given pixel of the curve (which corresponds to one X-axis tick representing time span / graph area in pixels )  3 values are shown - minimal, average and maximal. When there are values in the log database for all points of curve in a given time scale, all three values will be the same ( and the curve will look "pure"). See more on this below.
   *(4) - semitransparent color draws a swing between min and max values that the selected register took on a given plot section.
-  *(5) - Line of averaged values. In the case when for a selected plot scale 1 minute or more will fit into 1 pixel of the screen, then the average values of the register and the zone (4) of the range of values for estimating the distribution of the minimum and maximum peaks will be displayed on the graph. This averaging is drawn using cached data from the database, corresponding to the current scale of the view (in the database, averaging is stored for 1, 2, 5, 15 and 60 minutes). This method allows you to optimize the work with graphs when viewing large amounts of data on a different scale, while maintaining the speed of work. See below the features of working with charts.
+  *(5) - line of averaged values.
-  *(6) - line of values, when current view time scale matches log speed of the parameter so that there is either exact value, or its average/min/max are very close to exact value. (see video below for explanation)
+  *(6) - line of (almost) exact values.
-  *(7) - viewport slider. Changing its size with the help of the left and right sliders, you can move the viewport along the time axis, thus choosing the interval of interest with the same time scale
+  *(7) - viewport slider. You can move the viewport slider along the time axis, thus choosing the interval of interest with the same time scale
-  *(8) - viewport slider left handle
+  *(8) - viewport slider changing size left handle
-  *(9) - viewport slider right handle
+  *(9) - viewport slider changing size right handle
-Below is the video clearing how the curve's 4, 5, 6 parts are formed as well as min, max and avg values. In this example, the register's property was set to use data from log and save on every change. The sinus value was calcualted in a script executed each scan. Then system scan time was changed from 50 ms to 1 seconds. Becuase graph viewing area resoluon is 1 seconds, more data saved in database log for any given point when scan was 50 ms and that was highlighted with semitransparent zone along the curve.
-{{more_graph_data_in_one_sec.gif}}
+   See explanation below for the spots #4, #5, #6 .
 ==== Navigation in the graph viewing area ====
@@ Line 39: / Line 40: @@
 See the video:
-{{graph_navigate.gif?direct}}
+{{network:graph_navigate.gif?direct}}
+==== Graph curve display features ====
+Below is the video clearing how the curve's 4, 5, 6 parts are formed as well as min, max and avg values.
+In this example, the register's property was set to use data from log and save on every change. The sinus value was calcualted in a script executed each scan. Then system scan time was changed from 50 ms to 1 seconds. Becuase graph viewing area resoluon is 1 seconds, more data saved in database log for any given point when scan was 50 ms and that was highlighted with semitransparent zone along the curve.
+{{network:more_graph_data_in_one_sec.gif}}
+In the case when for a selected plot scale 1 minute or more will fit into 1 pixel of the screen, then the average values (5) of the register and the zone (4) of the range of values for estimating the distribution of the minimum and maximum peaks will be displayed on the graph. This averaging is drawn using cached data from the database, corresponding to the current scale of the view (in the database, averaging is stored for 1, 2, 5, 15 and 60 minutes). This method allows you to optimize the work with graphs when viewing large amounts of data on a different scale, while maintaining the speed of work.
+When the current view time scale matches log speed of the parameter, the (6) solid lined is drawn so that there is either exact value, or its average / min / max are very close to each other.
-==== Log data size considerations ====
-The log depth can be adjusted in the Setup / Settings menu:
-{{::log_depth_scr_shot.png?direct}}
-Theoretically, the volume of the archive is limited only by the size of the flash card.
-In practice, when the archive size is increased by more than 300MB, the 'response time' in the interface when working with WebHMI can increase and affect the comfort of the user's work. It should be said that storing large amounts of data is not the target application of the device, although it allows it to do so. These graphs should be used during the research stage of the monitoring object, keep the archive for analysis of certain situations (which after use can be deleted), etc. For long-term storage of a large amount of monitoring data, it is recommended to use a special server - fetching log data via the API or the Level2 cloud server option.