Select-A-Graph and yesterday.ini errors

[lidar37]

Select-A-Graph consistently shows wrong start time in build 1001. This also occurred in build 998. You have to set the time 5 minutes behind where you want the graph to start. For instance, the attached graph time selection begins at 1615 but the actual graph time is 1620.

Also note that the high temp for yesterday occurs at 1655 on the graph but occurs at 1625 in the attached yesterday.ini file.

[img][/img]

[steve]

Select-A-Graph consistently shows wrong start time in build 1001. This also occurred in build 998. You have to set the time 5 minutes behind where you want the graph to start. For instance, the attached graph time selection begins at 1615 but the actual graph time is 1620.

It's because the time selector component adds random seconds on to the time, so the first entry that Cumulus finds greater than or equal to that time is the entry after it. I'll see what I can do.

Also note that the high temp for yesterday occurs at 1655 on the graph but occurs at 1625 in the attached yesterday.ini file.

The entries in the monthly logs are snapshots of the values at the time of the entry. They don't necessarily capture the highs and lows.

[lidar37]

Select-A-Graph consistently shows wrong start time in build 1001. This also occurred in build 998. You have to set the time 5 minutes behind where you want the graph to start. For instance, the attached graph time selection begins at 1615 but the actual graph time is 1620.

It's because the time selector component adds random seconds on to the time, so the first entry that Cumulus finds greater than or equal to that time is the entry after it. I'll see what I can do.

Also note that the high temp for yesterday occurs at 1655 on the graph but occurs at 1625 in the attached yesterday.ini file.

The entries in the monthly logs are snapshots of the values at the time of the entry. They don't necessarily capture the highs and lows.

Ok, I'm confused. Does the yesterday.ini file pull data from the monthly.ini to get it's max/min values? If not, then what do you mean by your last comment?

[mcrossley]

The max/min are recorded in real time by Cumulus (assuming it is running and connected to a station!) and recorded in today.ini, at midnight the today.ini values get rolled over into yesterday.ini. If they exceed any existing records then they will also be recorded in the appropriate files.

The log files are snapshots of whatever the value is at the time of the snapshot. So they will not capture the highs/lows unless those values are sustained over the period of a snapshot time.

[lidar37]

The max/min are recorded in real time by Cumulus (assuming it is running and connected to a station!) and recorded in today.ini, at midnight the today.ini values get rolled over into yesterday.ini. If they exceed any existing records then they will also be recorded in the appropriate files.

The log files are snapshots of whatever the value is at the time of the snapshot. So they will not capture the highs/lows unless those values are sustained over the period of a snapshot time.

Where does data from Select-A-Graph come from, monthly.ini? If so, then the previous statements make sense. Thanks.

[mcrossley]

The graph data comes from the monthly log files Jul11log.txt etc. That is the only place data of that detail is kept.

[steve]

Perhaps the confusion came from my use of the term 'monthly logs', if so - apologies. It's a term I often use to describe the main detailed one-per-month Cumulus log files that Mark correctly refers to. I didn't mean the month.ini file, which just has a set of highs and lows for the current month in it.