2f:T1c42,function getHeatRatios(csvData, options = {}, constants = {}) { const { heatRatioWindowStart = 75000, heatRatioWindowEnd = 95000 } = constants; const { downstreamTemperatureSensor = (d) => +d.temp2, upstreamTemperatureSensor = (d) => +d.temp1, referenceTimer = (d) => +d.millisSinceReferenceTemp, heatPulseTimer = (d) => +d.millisSinceHeatPulse, timestamp = (d) => +d.rtcUnixTimestamp, // expects unix timestamps in seconds id = (d) => d.nodeId, downstreamTempDifference = (d) => +d.downstreamTempDifference, upstreamTempDifference = (d) => +d.upstreamTempDifference, filters = { temperature: (datum) => upstreamTemperatureSensor(datum) >= 0 && downstreamTemperatureSensor(datum) >= 0 && Math.abs(upstreamTemperatureSensor(datum)) < 40 && Math.abs(downstreamTemperatureSensor(datum)), heatRatioCalculation: (datum) => referenceTimer(datum) > 10000 && heatPulseTimer(datum) > heatRatioWindowStart && heatPulseTimer(datum) < heatRatioWindowEnd && // make sure the temperature hasn't dropped back down to the baseline reference temp downstreamTempDifference(datum) > 0 && upstreamTempDifference(datum) > 0 } } = options; csvData.sort((a, b) => { if (timestamp(+a) > timestamp(+b)) return 1; if (timestamp(+a) < timestamp(+b)) return -1; if (heatPulseTimer(+a) > heatPulseTimer(+b)) return 1; if (heatPulseTimer(+a) < heatPulseTimer(+b)) return -1; }); const dataGroupedById = csvData .filter(filters.temperature) .reduce((acc, curr) => { const rowToAdd = { ...curr, date: new Date(timestamp(curr) * 1000) }; if (!acc[id(curr)]) { acc[id(curr)] = [rowToAdd]; } else { acc[id(curr)].push(rowToAdd); } return acc; }, {}); const ids = Object.keys(dataGroupedById); const dataGroupedByHeatPulse = Object.fromEntries( Object.entries(dataGroupedById).map(([id, rawData]) => { let indexOfCurrentHeatPulse = 0; let dataBinnedByHeatPulse = []; for (let [index, datum] of rawData.entries()) { if (index === 0) { dataBinnedByHeatPulse[0] = [datum]; } const prevIndex = index - 1; if (prevIndex > 0) { // Conditions for determining a new heat pulse bucket should be created: // - if the millisSinceHeatPulse of the current datum is less than the millisSinceHeatPulse of the // previous datum // - if there was a data interruption (i.e., the unix_timestamp of the current datum is >3 seconds // apart from the unix_timestamp of the previous datum), count these as distinct heat pulse buckets if (referenceTimer(datum) < referenceTimer(rawData[prevIndex])) { indexOfCurrentHeatPulse++; dataBinnedByHeatPulse[indexOfCurrentHeatPulse] = [datum]; } else { dataBinnedByHeatPulse[indexOfCurrentHeatPulse].push(datum); } } } dataBinnedByHeatPulse = dataBinnedByHeatPulse.map((heatPulseData) => { const referenceTempData = heatPulseData.filter( (datum) => referenceTimer(datum) < 10000 ); const referenceDownstreamTemp = getMean( referenceTempData.map((datum) => downstreamTemperatureSensor(datum)) ); const referenceUpstreamTemp = getMean( referenceTempData.map((datum) => upstreamTemperatureSensor(datum)) ); const upstreamRollingMax = identifyRollingMax({ data: heatPulseData, accessor: upstreamTemperatureSensor }); const downstreamRollingMax = identifyRollingMax({ data: heatPulseData, accessor: downstreamTemperatureSensor }); const indexOfRollingMaxStart = Math.max( upstreamRollingMax, downstreamRollingMax ); const timeOfRollingMaxStart = heatPulseTimer( heatPulseData[indexOfRollingMaxStart] ); const upstreamLineGenerator = ss.linearRegressionLine( ss.linearRegression( heatPulseData .filter((d, i) => i >= rollingMaxStart) .map((d) => [heatPulseTimer(d), upstreamTemperatureSensor(d)]) ) ); const downstreamLineGenerator = ss.linearRegressionLine( ss.linearRegression( heatPulseData .filter((d, i) => i >= rollingMaxStart) .map((d) => [heatPulseTimer(d), downstreamTemperatureSensor(d)]) ) ); return heatPulseData.map((datum, i) => { const downstreamTempDifference = downstreamTemperatureSensor(datum) - referenceDownstreamTemp; const upstreamTempDifference = upstreamTemperatureSensor(datum) - referenceUpstreamTemp; const heatRatio = downstreamTempDifference / upstreamTempDifference; const modeledDownstreamTempDifference = i < indexOfRollingMaxStart ? undefined : downstreamLineGenerator(heatPulseTimer(datum)) - referenceDownstreamTemp; const modeledUpstreamTempDifference = i < indexOfRollingMaxStart ? undefined : upstreamLineGenerator(heatPulseTimer(datum)) - referenceUpstreamTemp; const modeledHeatRatio = modeledDownstreamTempDifference === undefined ? undefined : modeledDownstreamTempDifference / modeledUpstreamTempDifference; return { ...datum, referenceDownstreamTemp, downstreamTempDifference, referenceUpstreamTemp, upstreamTempDifference, heatRatio, referenceTempData, modeledDownstreamTempDifference, modeledUpstreamTempDifference, modeledHeatRatio }; }); }); return [id, dataBinnedByHeatPulse]; }) ); const meanHeatRatiosGroupedById = Object.fromEntries( Object.entries(dataGroupedByHeatPulse).map(([nodeId, heatPulses]) => { const meanHeatRatios = heatPulses.map((heatPulseData) => { const timestampForMeanHeatRatio = timestamp(heatPulseData[0]) - Math.round(referenceTimer(heatPulseData[0]) / 1000 + 10); const nodeID = id(heatPulseData[0]); const date = new Date(timestampForMeanHeatRatio * 1000); const heatPulseDataUsedForCalculatingHr = heatPulseData.filter( filters.heatRatioCalculation ); const meanHeatRatio = getMean( heatPulseDataUsedForCalculatingHr.map((datum) => datum.heatRatio) ); const modeledMeanHeatRatio = getMean( heatPulseData .filter((d) => d.modeledHeatRatio !== undefined) .map((d) => d.modeledHeatRatio) ); return { date, nodeID, meanHeatRatio, modeledMeanHeatRatio, heatPulseDataUsedForCalculatingHr, heatPulseData }; }); return [nodeId, meanHeatRatios]; }) ); return { ids, data: meanHeatRatiosGroupedById }; }30:T447,Plot.plot({ y: { grid: true }, x: { grid: true }, marks: [ Plot.dot(poinsettia_hp_2, { x: "millisSinceHeatPulse", // x: "millisSinceReferenceTemp", y: "temp1", stroke: (d, i) => { const maxAverageWindowStart = rollingMaxStart; return i >= maxAverageWindowStart && i < maxAverageWindowStart + 10 ? "orange" : "green"; } }), Plot.dot(poinsettia_hp_2, { x: "millisSinceHeatPulse", // x: "millisSinceReferenceTemp", y: "temp2", stroke: (d, i) => { const maxAverageWindowStart = rollingMaxStart; return i >= maxAverageWindowStart && i < maxAverageWindowStart + 10 ? "red" : "steelblue"; } }), Plot.line( [ [65000, temp1LineGenerator(65000)], [93000, temp1LineGenerator(93000)] ], { stroke: "green" } ), Plot.line( [ [65000, temp2LineGenerator(65000)], [93000, temp2LineGenerator(93000)] ], { stroke: "steelblue" } ) ] })31:[{"id":"715db1223e067f00500780077febc6cebbdd90c151d3d78317c802732252052ab0e367039872ab9c77d6ef99e5f55a0724b35ddc898a1c99cb14c31a379af80a","href":"https://static.observableusercontent.com/files/715db1223e067f00500780077febc6cebbdd90c151d3d78317c802732252052ab0e367039872ab9c77d6ef99e5f55a0724b35ddc898a1c99cb14c31a379af80a","name":"penguins.csv","type":"text/csv","lastModified":1678159517151.194,"size":13506},{"id":"fd8073528ed2875bfa6d34928ba7e757ec5cf6b85616c1ac5198f5afa4ab434aa73e72febe89a29978f209d271e3ba8177febb2be745494de4a9551fbb3170ae","href":"https://static.observableusercontent.com/files/fd8073528ed2875bfa6d34928ba7e757ec5cf6b85616c1ac5198f5afa4ab434aa73e72febe89a29978f209d271e3ba8177febb2be745494de4a9551fbb3170ae","name":"DATALOG@1.CSV","type":"text/csv","lastModified":1678159848640.66,"size":35019470},{"id":"1decb40ea14951a0e5c260e79a687275a65ab5a9d50caf9932a3d0ee13debec985a5394812e3d59c3aa28da94197f34b8bd938bebbc8ddd51259a4ace6743de8","href":"https://static.observableusercontent.com/files/1decb40ea14951a0e5c260e79a687275a65ab5a9d50caf9932a3d0ee13debec985a5394812e3d59c3aa28da94197f34b8bd938bebbc8ddd51259a4ace6743de8","name":"datalog_2023-04-03.parquet","type":"application/octet-stream","lastModified":1681176618470.419,"size":1231458}] 2d:{"title":"Plot Scatterplot","cells":[{"id":0,"value":"# Plot Scatterplot\n\nUsing [Plot.dot](https://observablehq.com/@observablehq/plot-dot).","pinned":false,"hidden":false,"mode":"md"},{"id":4,"value":"data = FileAttachment(\"penguins.csv\").csv({typed: true})","pinned":true,"hidden":false,"mode":"ojs"},{"id":6,"value":"Plot.plot({\n inset: 8,\n grid: true,\n color: {\n legend: true,\n },\n marks: [\n Plot.dot(data, {x: \"flipper_length_mm\", y: \"body_mass_g\", stroke: \"sex\"})\n ]\n})","pinned":true,"hidden":false,"mode":"ojs"},{"id":58,"value":"SELECT *\nFROM \"DATALOG@1.CSV\"\nWHERE cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) > '2023-02-22'\n AND cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) < '2023-02-24'\nLIMIT 100","pinned":true,"hidden":false,"mode":"sql","output":"hp1","database":"DATALOG@1.CSV"},{"id":73,"value":"Plot.plot({\n y: {\n grid: true\n },\n x: {\n grid: true\n },\n marks: [\n Plot.dot(hp1, {\n x: \"millisSinceHeatPulse\",\n y: \"temp1\",\n stroke: \"green\"\n }),\n Plot.dot(hp1, {\n x: \"millisSinceHeatPulse\",\n y: \"temp2\",\n stroke: \"steelblue\"\n })\n ]\n})","pinned":true,"hidden":false,"mode":"ojs"},{"id":76,"value":"datapointsForHeatRatioCalculation = hp1.filter(\n (d) => d.millisSinceReferenceTemp > 80000\n)","pinned":true,"hidden":false,"mode":"ojs"},{"id":80,"value":"SELECT *\nFROM \"DATALOG@1.CSV\"\nWHERE cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) > '2023-02-21'\n AND cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) < '2023-02-24'\nLIMIT 100","pinned":true,"hidden":false,"mode":"sql","output":"bambooHeatPulses","database":"DATALOG@1.CSV"},{"id":92,"value":"bambooHeatRatios = getHeatRatios(bambooHeatPulses)","pinned":true,"hidden":false,"mode":"ojs"},{"id":98,"value":"Plot.plot({\n x: {\n type: \"time\"\n },\n marks: [\n Plot.line(bambooHeatRatios.data[\"sapflow-dotmote-labs-33\"], {\n x: \"date\",\n y: \"modeledMeanHeatRatio\"\n })\n ]\n})","pinned":true,"hidden":false,"mode":"ojs"},{"id":90,"value":"$2f","pinned":true,"hidden":false,"mode":"ojs"},{"id":88,"value":"function getMean(numbers) {\n if (!Array.isArray(numbers)) {\n numbers = Array.from(arguments);\n }\n\n if (!numbers.length) {\n return undefined;\n }\n\n return numbers.reduce((acc, curr) => +acc + +curr, []) / numbers.length;\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":112,"value":"SELECT *\nFROM \"DATALOG@1.CSV\"\nWHERE cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) > '2023-02-03'\n AND cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) < '2023-02-06'\nLIMIT 100","pinned":true,"hidden":false,"mode":"sql","output":"poinsettiaHp","database":"DATALOG@1.CSV"},{"id":117,"value":"Plot.plot({\n y: {\n grid: true\n },\n x: {\n grid: true\n },\n marks: [\n Plot.dot(poinsettiaHp, {\n x: \"millisSinceHeatPulse\",\n // x: \"millisSinceReferenceTemp\",\n y: \"temp1\",\n stroke: \"green\"\n }),\n Plot.dot(poinsettiaHp, {\n x: \"millisSinceHeatPulse\",\n // x: \"millisSinceReferenceTemp\",\n y: \"temp2\",\n stroke: \"steelblue\"\n })\n ]\n})","pinned":true,"hidden":false,"mode":"ojs"},{"id":128,"value":"SELECT *\nFROM \"DATALOG@1.CSV\"\nWHERE cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) > '2023-02-02'\n AND cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) < '2023-02-10'\nLIMIT 100","pinned":true,"hidden":false,"mode":"sql","output":"poinsettiaHeatPulses","database":"DATALOG@1.CSV"},{"id":133,"value":"poinsettiaHeatRatios = getHeatRatios(\n poinsettiaHeatPulses,\n {},\n { heatRatioWindowStart: 75000, heatRatioWindowEnd: 95000 }\n)","pinned":true,"hidden":false,"mode":"ojs"},{"id":135,"value":"Plot.plot({\n x: {\n type: \"time\"\n },\n marks: [\n Plot.line(\n poinsettiaHeatRatios.data[\"sapflow-dotmote-labs-33\"].slice(225, 300),\n {\n x: \"date\",\n y: \"modeledMeanHeatRatio\"\n }\n )\n ]\n})","pinned":true,"hidden":false,"mode":"ojs"},{"id":163,"value":"SELECT *\nFROM \"datalog_2023-04-03.parquet\"\nWHERE cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) > '2023-03-31'\n AND cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) < '2023-04-01'\nLIMIT 100","pinned":true,"hidden":false,"mode":"sql","output":"poinsettia_hp_2","database":"datalog_2023-04-03.parquet"},{"id":229,"value":"SELECT *\nFROM \"datalog_2023-04-03.parquet\"\nWHERE cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) > '2023-03-31'\n AND cast(to_timestamp(rtcUnixTimestamp) as TIMESTAMP) < '2023-04-01'\nLIMIT 1000","pinned":true,"hidden":false,"mode":"sql","output":"poinsettiaHeatPulses2","database":"datalog_2023-04-03.parquet"},{"id":180,"value":"/*\nReturns the first index of the max N window\n*/\n\nfunction identifyRollingMax({ data, n = 10, accessor = (d) => d.value }) {\n let windowStart = 0;\n let windowEnd = n;\n let maxAverageValue = getMean(\n data.slice(windowStart, windowEnd).map(accessor)\n );\n let maxAverageValueWindowStart = 0;\n\n if (n > data.length) {\n return undefined;\n }\n\n let potentialNewMaxAverageValue;\n while (windowEnd < data.length) {\n windowStart++;\n windowEnd++;\n potentialNewMaxAverageValue = getMean(\n data.slice(windowStart, windowEnd).map(accessor)\n );\n if (potentialNewMaxAverageValue >= maxAverageValue) {\n maxAverageValueWindowStart = windowStart;\n maxAverageValue = potentialNewMaxAverageValue;\n }\n }\n return maxAverageValueWindowStart;\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":182,"value":"identifyRollingMax({ data: poinsettia_hp_2, accessor: (d) => d.temp1 })","pinned":true,"hidden":false,"mode":"ojs"},{"id":184,"value":"identifyRollingMax({ data: poinsettia_hp_2, accessor: (d) => d.temp2 })","pinned":true,"hidden":false,"mode":"ojs"},{"id":189,"value":"rollingMaxStart = Math.max(\n identifyRollingMax({ data: poinsettia_hp_2, accessor: (d) => d.temp1 }),\n identifyRollingMax({ data: poinsettia_hp_2, accessor: (d) => d.temp2 })\n)","pinned":true,"hidden":false,"mode":"ojs"},{"id":196,"value":"ss.linearRegression(\n poinsettia_hp_2\n .filter((d, i) => i >= rollingMaxStart)\n .map((d) => [d.millisSinceHeatPulse, d.temp1])\n)","pinned":true,"hidden":false,"mode":"ojs"},{"id":202,"value":"temp1LineGenerator = ss.linearRegressionLine(\n ss.linearRegression(\n poinsettia_hp_2\n .filter((d, i) => i >= rollingMaxStart)\n .map((d) => [d.millisSinceHeatPulse, d.temp1])\n )\n)","pinned":true,"hidden":false,"mode":"ojs"},{"id":212,"value":"temp2LineGenerator = ss.linearRegressionLine(\n ss.linearRegression(\n poinsettia_hp_2\n .filter((d, i) => i >= rollingMaxStart)\n .map((d) => [d.millisSinceHeatPulse, d.temp2])\n )\n)","pinned":true,"hidden":false,"mode":"ojs"},{"id":171,"value":"$30","pinned":true,"hidden":false,"mode":"ojs"},{"id":194,"value":"ss = require(\"simple-statistics\")","pinned":true,"hidden":false,"mode":"ojs"}],"files":"$W31","stdlib":"1","titling":"auto"}