The time series, something you see a lot more of once you get into industry and then never get away from. Great for telling a story and finding out causation. Below we will share what we think its good for, how to get more out of it, and where we don’t often find any value in using it.
This is article 4/7 in our series focussing on the bread and butter visualisations our analytics team use.
To receive a copy of the whole series now go to www.clarofy.ai/download
Critical in exploring the data and significantly useful in evaluation; insight gained from visualisations inform the whole analytical workflow. Some types of visualisations lend themselves better to one use over the other and, making the decision of which to use in each application is a learning process. At Interlate, our experience with analytics in Minerals processing plants has given us an appreciation for how to make these decisions, and we will outline some of our insights in the following series.
- Justifiably, looking for changes in variables over time.
- Ensuring that both long term and short-term trends are examined; looking at changes over the months or years, analysis related to control at a small time periods such as 24 or 48 hours.
- When stacked or related measures are used, you can find how relationships between variables have changed over time. – however sometimes this is more easily seen in scatter plots coloured by time.
- Great for identifying how much lag you might need to apply to certain variables to see the relationships on scatter plots.
- Identifying trials, modes or events in plant operation.
- Finding relationships between variables
- Noisy data that fluctuates heavily within a day or hour and could obscure valuable relationships.
- High data density as it exacerbates issues found in fluctuations.
Tips & Tricks:
- Aggregate the data when dealing with variables that fluctuate heavily, however be careful of introducing or removing relationships when aggregating
- Try different methods of aggregation, min, max, percentiles, average, moving averages.
- Be open to having a look at any available categorical variables over time, in addition to the numeric variables. Being critical of an operation mode may be unfair if 90% of its data occurred before a particular upgrade or with a certain ore type.
- Spend some time looking at some time series if you are unfamiliar with the plant. It gives important contextual information as to the stability of the plant, how often shutdowns occur, when different modes in the plant are in operation.
- After transforming data (such as integrating feed grade out of recovery, or modelling a dependent variable) compare it again on a timescale to see if there are errors or different correlations with time.
This article series will focus on our key visuals, Interlate hope to share our experience to others and provide a robust understanding for their place in Clarofy our visualisation and data analytics application. No software installation required and runs straight from your browser www.clarofy.ai