2c:null 2d:["$","title",null,{"children":"Energy Worksheet Example | Joshua Cooper | Observable"}] 32:T5b0,md`Off peak is defined as slots 0-14 or midnight (00:00) up to and including the 07:00 to 07:30 period. Evening is 16:00 to 23:30 or slot 32 to 47 - so it includes that final half hour at the end of the day. Day time is slot 15 through and including 31 or 07:30 to 15:30. This of course is somewhat arbitrary, but has been used in behavioural research. Week day and weekend are self explainatory. Baseline is interesting and has some different interpretations in literature. For our working purposes we consider it the non-zero minimum for a day. Zero is possible and we do see it in homes with battery or solar, remember we are looking at imported or purchased energy in this data stream. The rationale is that at some point the in day baseload will come through as a background; less true as the home is more self sufficient in producing or storing energy. The statistics that are calculated are on a weekly basis, hence the 53 items in the array. The index of the array is the week number according to a standard calendar function for a year, MINUS 1 ℺ as it is an index. For instance Week 1 or index 0 for year 2020 is 30 Dec 2019 to 5 Jan 2020. A good resource for explaining these features extracted from the energy signal are found in Peter Grindrod, Mathematical Underpinnings of Analytics, (2014) OUP, Oxford. ISBN: 9780198725091 page 135. BTW, Peter if you are reading this, I really like your book! `33:[] 30:{"id":"a755b0939c3aeac3","slug":"energy-worksheet-example","title":"Energy Worksheet Example","thumbnail":"73e7740373d952ae836714adf9ef26e674c4494dbdb0f9056841bc02e11c48cd","access_level":"public","create_time":"$D2021-02-07T11:42:04.691Z","update_time":"$D2022-03-02T15:09:06.279Z","view_time":null,"first_public_time":"$D2021-02-26T10:07:17.632Z","has_unpublished_edits":null,"fork":false,"stars":1,"starred":false,"owner":{"type":"team","id":"20e510411e88c2f1","login":"joshuacooper","name":"Joshua Cooper","avatar_url":"https://avatars-next.observableusercontent.com/814f458da4ea4bcfddd1fc64693fcb58da584f45f67098626643f24a0432a1c4","tier":"starter_2024"},"version":673,"publish_version":673,"role":null,"authored":false,"authors":[{"type":"individual","id":"5514a468c7245fb1","login":"joshuacooper","name":"Joshua Cooper","avatar_url":"https://avatars-next.observableusercontent.com/814f458da4ea4bcfddd1fc64693fcb58da584f45f67098626643f24a0432a1c4","tier":null}],"members":[{"type":"individual","id":"5514a468c7245fb1","login":"joshuacooper","name":"Joshua Cooper","avatar_url":"https://avatars-next.observableusercontent.com/814f458da4ea4bcfddd1fc64693fcb58da584f45f67098626643f24a0432a1c4","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":2,"collection":{"id":"04adf3cf439dcfe5","slug":"smart-tariff-smart-comparison","title":"Smart Tariff Smart Comparison","owner":{"id":"20e510411e88c2f1","login":"joshuacooper","name":"Joshua Cooper"}},"type":"notebook","body":{"titling":"auto","title":"Energy Worksheet Example","cells":[{"id":0,"value":"md`# Energy Worksheet Example\nThis worksheet shows the test data being loaded and then a simple (if not useless!) plot of the 30 minute data for the year.\nNormally you would use a public token or login through an STSC object to get a session token, then load your data.\nThe energy data can be aggregated, analysed etc.\n\nNode.js examples are found here https://runkit.com/ijosh/stsc-energy-example/0.4.4\nalong with the source code for the library\nhttps://bitbucket.org/LDNSFO/stsc-tsc/\n\nNPM is also a good place to pick up and use. The documentation should be found there as well as the git repository. https://www.npmjs.com/package/stsc-tsc\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":316,"value":"vl.markLine()\n .data(myenergy.data)\n .encode(\n vl.x().fieldT([0]),\n vl.y().fieldQ([1])\n )\n .render()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":468,"value":"md`There are methods - or function calls - that can be made to your energy object such as load, profile, consumption, addEnergy, shiftEnergy and weekly. And as Energy is a Timeseries, you can use the lower level aggregation functions to manipulate time range and aggregations of the Energy object.\n\nNotice in the example byHHSlotDOW() below, it is acutally a method on the Timeseries, making the Energy object more powerful. The design principle is Energy should have context beyond a simple series of numbers (a Timeseries). Enerergy will be expressing units of kWh, have measures like profile that are specific in classifying energy use; whereas Timeseries functions are general and manipulate data.\n\nA classic example is the load function. We load Energy into a Timeseries because the raw storage for Energy, the file that gets downloaded, is specifically representing Imported energy - no negative values for example. It has a fixed interval of 30 minutes. It means we can be efficient in how that data is transmitted.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":290,"value":"md`### Summarising energy data\nIt is useful to see the half hourly patterns through out different days. Built in to the energy class is a summarisation method that shows the average consumption by 30 minute period and then broken down for each day.\n\nThe resulting data is an array of array of numbers with index 0-6 representing Monday (day 0) to Sunday (day 6) and 48 half hourly time slots. The time runs from midnight 00:00 for slot 0 to 23:30 for slot 47. Slot 0 contains all of the kWh consumed for the period of 00:00 to 00:30.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":417,"value":"sumHHForDayOfWeek = myenergy.byHHSlotDOW()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":520,"value":"table(sumHHForDayOfWeek)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":484,"value":"md`The time slots are then aggregated summing across the days that are in the entire data set. For instance, all Mondays (first row) from 00:00 to 00:30 (first column) give me the total kWh. Other aggregrate functions can be used by passing a parameter into the \n\\`\\`\\`Javascript \nbyHHSlotDOW({aggregate: my.Timeseries.AGG_MEAN})\n\\`\\`\\`\nThe various aggregate functions are:\n\\`\\`\\`Javascript \nmy.Timeseries.AGG_SUM\nmy.Timeseries.AGG_COUNT\nmy.Timeseries.AGG_VAR\nmy.Timeseries.AGG_STDDEV\nmy.Timeseries.AGG_MEAN\nmy.Timeseries.AGG_MAX\nmy.Timeseries.AGG_MIN\nmy.Timeseries.AGG_MEDIAN\n\\`\\`\\`\n\nThe next table shows the AGG_MEAN function being used as a parameter - which is a lot more useful!\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":488,"value":"table(myenergy.byHHSlotDOW({aggregate: my.Timeseries.AGG_MEAN}))","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":562,"value":"md`## Behavioural features\n\nSeveral analysis methods use statistics calculated on a weekly basis looking daily averages, various time ranges during the day, etc. \n\nWe have built in a behavioural profile to help week on week trending as well as be able to make comparisons between households or larger \"average\" household characteristics using something like machine learning or statistical clustering.\n\nEach row below represents a week.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":356,"value":"{\n let weeklystats = myenergy.byWeek();\n console.log(weeklystats);\n let prettyweeklystats = [];\n weeklystats.forEach((item, i) => {\n prettyweeklystats.push({ weeknumber: i, ...item });\n });\n return table(prettyweeklystats)\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":478,"value":"$32","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":587,"value":"vl.markLine()\n .title(\"Histogram of readings\")\n .data(freq)\n .encode(\n vl.x().fieldQ([0]).title(\"Wh\"),\n vl.y().fieldQ([1]).title(\"count\")\n )\n .render()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":582,"value":"freq = { \n let reshapeData = [];\n let hist = myenergy.histogram();\n hist.forEach((value, index) => {\n reshapeData.push([index, value]);\n });\n return reshapeData;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":593,"value":"md`A histogram function is built in, note the index of the array is the value being counted. There is a frequency plot above showing a concentration of use around the 300-400 watt regions.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":620,"value":"vl.markBar()\n .title(\"Consumption by month\")\n \n .data(myenergy.byMonth())\n .encode(\n vl.x().fieldT([0]).title(\"2020 - 2021\"),\n vl.y().fieldQ([1]).title(\"Wh\")\n )\n .render()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":631,"value":"md`There is a handy monthly calculator that will sum the energy from within the calendar month. This will probably change in the future to be a function rather than a property, so that different aggregation functions, numbers of months and a few more parameters can be passed.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":635,"value":"md`### Further examples\n\nFollowing on, the next topic is tariffs which can be found here at https://observablehq.com/@joshuacooper/tariff-calculations`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":254,"value":"myenergy = {\n let energy = new my.Energy();\n energy._token = 'resource_test';\n await energy.load();\n return energy;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":284,"value":"my = require('https://bundle.run/stsc-tsc@1.0.13');","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":306,"value":"import {vl} from '@vega/vega-lite-api'","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":517,"value":"import {table} from '@tmcw/tables/2'","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W33","stdlib":"1"}} 31:"$30:body"