Changes in openair 0.9-0 [2013-02-13] ===================================== * Fix bug introduced in 0.8-0 for scatterPlot surfaces * Add statistic and percentile option to polarAnnulus; allow Conditional Probability Functions * Allow percentile intervals to be considered in polarPlot when statistic = "cpf" * Correct calculation of AOT40 in aqStats to take account of daylight hours and growing season (Apr. to Sep.). Can use latitude/longitude, if supplied * Fix bug in aqStats when non predefined pollutant used * Allow log10 factor of 2 lines in scatterPlot when x and/or y are on a log scale * Allow users to supply own transform/inverse functions to hexbin * Don't try and pad-out missing data in rollingMean when returning results * Make latticeExtra a 'depends' * Don't copy-down data by default in timeAverage when requested averaging time is < original. Better handling when 'site' is present * Update AURN help file site information; add site type * Show where data removed when min.bin > 1 for polarPlot * Add 'trim' value to percentiles in polarPlot to extract more source information. Changes in openair 0.8-0 [2012-11-30] ===================================== * make sure missing dates are plotted properly in scatterPlot * fix regression in corPlot - main did not work * remove time zone options in import - users must supply data in GMT (UTC). Too many problems introduced due to daylight saving time * Use Legates and McCabe Coefficient of Efficiency in modStats in place of the Index of Agreement - easier to interpret. * Allow type = "month" etc to be used in timeVariation (was variable clash). Allow more flexibility when group and type are used * Better handling of user-defined limits in polarPlot and scatterPlot (method = "level") * Add optional mean line option to percentileRose. * Correct pollutionRose documentation about comparing 2 data sets (first subtracted from second) * Fix bug in timeVariation that showed extra NA level for certain groups/types * Fix scaling bug in polarCluster when there is negative data e.g. x = "temp" * Allow statistic = "median" in trajLevel * NEW FUNCTION timeProp to plot time series by category as a barchart * Fix windRose bug when wind direction name was not wd * Fix bug in importAURN when pollutant = "all" was specified * Allow minimum value of breaks in ws to be above minimum ws value, but warn. * Allow day to be numeric in selectByDate to select days of the month * Better base maps with trajectory plotting (filled and alpha transparency) * Fix windRose bug where all data are missing * Use higher resolution mapdata not maps package for trajectory plots * Allow method = "hexbin" in trajLevel for hexagonal binning of trajectory frequencies * Refine rollingMean to allow moving window to be aligned centre/center, left or right + better treatment of ends when data capture threshold used. Use option 'width' rather than 'hours' because non-hourly data can be considered. * Fix pollutionRose scaling issue, which sometimes missed the lower interval * Fix timeAverage data capture issue - was not always setting data to NA * Output clusters as C1, C2 etc. not 1, 2. * Use same labelling in calendarPlot as other functions (wrong month order in non-English locales) * Export dendogram in corPlot - see example in help file * Fix scaling bug in polarPlot when radial variable was negative and small * Update 'mydata' to set negative data to NA * Add method = CPF to percentile rose and polarPlot * Update trajectory analysis to allow Potential Source Contribution Function (PSCF) and Concentration Weighted Trajectory (CWT) to be considered Changes in openair 0.7-0 [2012-05-30] ===================================== * Allow more flexible layout when two or more pollutants are used with polarPlot and type is default * Fix colour scale problem for windRose when interval does not exist * Added 'nativeRaster' class handler for GoogleMapsPlots * (test) Fix for GoogleMapsPlot xlim, ylim * Do not remove missing data from timeVariation * Fix bug in FAC2 in modStats when observations = 0 (exclude from calculation because indeterminate) * Fix bug in windRose for empty panels and panel labelling when data missing * New option start.day in cutData that affects most plots; allows users to set start day when type = "weekday" * Allow categorical scales in calendarPlot * Allow two met data sets to be compared in pollutionRose * Allow statistic = "percentile" to be used with polarPlot * Allow 'method' to be passed to cor in corPlot * Added more back trajectory locations. * Try harder to plot polarPlots when there is insufficent data to calculate a smooth surface. * Rename some variables in timeAverage to avoid variable clash. * Add gridded frequency capability to trajLevel * Fix bug in colour scaling in polarFreq when statistic = "stdev" (0 would not be plotted) * Fix bug in summaryPlot: type = "density" broken in recent versions Changes in openair 0.6-0 [2012-04-10] ===================================== * Add y.relation option to timePlot * Fix interpolation bug in calcFno2 and names in documentation. * Refine conditionalQuantile scales * Provide volatile and non-volatile components for FDMS PM10 and PM2.5 in importKCL - now consistent with importAURN * NEW FUNCTION conditionalEval for model evaluation - allows other variable performance to be assessed. * Make lattice strips white by default for cleaner look on complicated plots * Complete re-write of import to simplify - changes are NOT backward compatible but will allow more developments * Allow line breaks in titles using \n to work woth quickText - thanks to Karl * Allow better annotation of calendarPlot - highlight values above/below a certain threshold. Changes in openair 0.5-25 [2012-03-16] ===================================== * Fix bug when type = "wd" - would add missing data to north sector * Add observed histogram to conditionalQuantile * Fix bug in timeAverage when ws was not available but wd was * Temporaryfix to GoogleMapsPlot documentation due to new package version Changes in openair 0.5-23 [2012-02-28] ===================================== * Fix bug in windRose when paddle = FALSE Changes in openair 0.5-22 [2012-01-12] ===================================== * Fix scaling bug that could sometime affect polarPlot grid lines * Fix regression in importKCL that was intruduced in 0.5-21 * Make sure wd data are rounded to 10 degrees in polarFreq * Fix date padding issue in smoothTrend when type = "site" * windRose now gives mean ws in each panel rather than count * add otion date.format to timePlot for more control over date format on axis Changes in openair 0.5-21 [2012-02-21] * Use C++ code for rolling mean calcs. Much faster, more to follow * NEW function trajCluster to carry out cluster analysis on back trajectories * Simple model ranking available in modStats Changes in openair 0.5-18 [2012-01-16] * Update trajectory files to 2011 and add Berlin, Paris * Add option seg to pollutionRose to control the width of the segments * Add option start.day to timeVariation to control the order of weekdays. * Fixed bug in polarAnnulus where 360 degree winds were absent. * Remove dependency on zoo and proto. * Allow importAURN to import new ws/wd from pre-calculated WRF data at AURN sites Changes in openair 0.5-17 [2012-01-12] ===================================== * Update to [IN DEVELOPMENT] GoogleMapsPlot. (1 lat,lon default handling) Changes in openair 0.5-15 [2011-12-08] ===================================== * Update imports etc. Changes in openair 0.5-15 [2011-12-08] ===================================== * Update to [IN DEVELOPMENT] GoogleMapsPlot. * Tidy up calcFno2 plotting. * Improve speed of timeAverage by removing date.pad. * Add vector averaging option to timeAverage for wind speed. Changes in openair 0.5-14 [2011-11-29] ===================================== * NEW FUNCTION polarCluster for undertaking k-means clustering of bivariate polar plots * Remove any considerations of time zones in selectByDate. If start and end are supplied, whole days based on Date format are used to select. Now accepts start/end in the form "YYYY-mm-dd" as well as UK format e.g. dd/mm/YYYY. Changes in openair 0.5-13 [2011-11-18] ===================================== * Major update to polarPlot allowing variables other than "ws" to be plotted with wind direction. * Improve documentation for polarPlot/TaylorDiagram * Update openair citation information to Journal article Changes in openair 0.5-12 [2011-11-14] ===================================== * do not clip polarPlot concentrations unless upper is supplied * new option 'intervals' for percentileRose * add min.bin option to polarAnnulus * add Index of Agreement to modStats * new averaging time "season" for timeAverage * better treatment of avg.time = "season" and type = "season" in TheilSen and smoothTrend * use bootstrap methods to calculate 95% confidence intervals in the mean for timeVariation * New option 'difference' in timeVariation to show difference between two variables with bootstrap 95% CI in the mean * Byte-compile package * Place key on right be default in scatterPlot to avoid clash with x or y axis labels. * Include all colour schemes defined in the RColorBrewer package Changes in openair 0.5-11 [2011-09-14] ===================================== * update to corPlot, scatterPlot, smoothTrend, linearRelation, percentileRose, trajPlot, trajLevel, timeVariation, TaylorDiagram, timePlot, summaryPlot improved ... handling * remove warnings when importing air pollution data * additional update to corPlot added pollutant option, and openair class output * tweak scaling on percentileRose deal with negative data * change MannKendall to use TheilSen for all estimates for consistency. May slightly affect some p estimates. Changes in openair 0.5-10 [2011-09-12] ===================================== * update to calendarPlot, kernelExceed, MannKendall and conditionalQuantile improved ... handling * minor update to GoogleMapsPlot Changes in openair 0.5-9 [2011-09-08] ===================================== * allow model performance change to be considered in TaylorDiagram * update to windRose and pollutionRose stat related annotation * update to polarPlot, ploarFreq and polarAnnulus improved ... handling Changes in openair 0.5-8 [2011-09-02] ===================================== * NEW FUNCTION importMeta to import site meta data from air pollution networks Changes in openair 0.5-7 [2011-08-30] ===================================== * update to trendLevel: improved ... handling * update to windRose/pollutionRose: added statistic option "abs.count"; improved scaling of segment widths; improved ... handling Changes in openair 0.5-6 [2011-08-30] ===================================== * NEW FUNCTIONS importTraj, trajPlot, trajLevel for importing and plotting pre-defined HYSPLIT back trajectories. These functions are under active development and are for testing purposes only! * Fix error in % upper/lower uncertainty intervals for MannKendall * More control over trend information placement for MannKendall - arguments text.col, lab.frac (for vertical position) and lab.cex (for font size) Changes in openair 0.5-4 [2011-08-04] ===================================== * Use roxygen2 for package documentation and future maintainability * Allow splitByDate to handle mutiple sites and output a new column controlled by argument 'name' * More options for MannKendall: control of colour and variable x-axis scales Changes in openair 0.5-0 [2011-08-04] ===================================== * NEW FUNCTION corPlot for correlation matrices * add sep argument to import Changes in openair 0.4-23 [2011-08-03] ===================================== * enhance 'normalise' option in timePlot * add annotate option to windRose/pollutionRose * NEW FUNCTION TaylorDiagram for model evaluation. Changes in openair 0.4-22 [2011-06-03] ===================================== * NOTE - use reshape2 in place of reshape for speed and reliability * allow more than one pollutant with percentileRose * fixed title bug with scatterPlot - not shown for some methods * modifed key handling on plots using drawOpenKey, so key = NULL or FALSE now removes colour key. * added method = "level" to scatterPlot for binning data with optional smoothing, plus other code clean-ups * fix bug in conditionalQuantile that labelled plots wrongly when site(s) has missing data * added ref.x and ref.y to timePlot to allow refernce lines to be added * do not remove calm wind speed conditions in any functions where this is unecessary Changes in openair 0.4-21 [2011-06-03] ===================================== * fixed bug in polarAnnulus that resulsted in a failure to annotate the plot properly with period = "trend" and less than 1 year of data; improved smoothing default options used * new 'statistic' option to pollutionRose to show contribution to counts and contribution to the mean. The latter is useful for displaying those wind directions that make most contribution to teh overall mean. Panel mean is also now shown. * small change to final example of re-shaping data in importKCL * timeAverage can now expand data to shorter time periods e.g. hourly to 15-minute. This makes it more flexible to combine data sets with differeing averging times. For example, daily mean particle data can be expanded to 1-hour means and combined with an hourly meteorological data set. * Fix treatment of calms in checkPrep, which affected cases with zero wind speeds in timeAverage * Fix to vector averaging in timeAverage (did not include wind speed in calculations). For most data this will make very little difference, but will be more important for low wind speeds and/or variable wind directions. * allow type = 'variable' in smoothTrend * Add avg.time to smoothTrend for "month" or "year" averages Changes in openair 0.4-20 [2011-05-31] ===================================== * Do not remove calms in timeAverage * NEW FUNCTION importSAQN to import data from the Scottish Air Quality Network Changes in openair 0.4-19 [2011-05-19] ===================================== * improved import date.name/time.name handling of spaces * Improved scaling for polarFreq * scatterPlot type = "wd" labels aligned * Added option "statistic" to polarPlot, which can now consider "mean", "median", "max" (maximum), "frequency". "stdev" (standard deviation) or "weighted.mean" in a similar way to polarFreq * Fixed bad.24 and international tz bugs in import * updated import/import.2 Changes in openair 0.4-18 [2011-05-19] ===================================== * Better handling of missing data in smoothTrend/MannKendall; particularly when there are mutiple sites * Do not run checkPreo on selectByDate because it removes calms * Fixed scaling bug in polarFreq when ws.int != 1 * use avg.time in calcPercentile instead of 'period' to ensure consistency with other functions (timePlot would fail with percentiles) * return NA in aqStats when no data present; previously it tried to calculate quanities, returning -Inf etc Changes in openair 0.4-17 [2011-04-21] ===================================== * use readRDS and not .readRDS (caused failure on R Dev) * openair now depends on >=R2.13.0 * NEW FUNCTION sqStats to calculate common and pollutant-specfic air quality statsitics by year and site. * Changed option data.capture in rollingMean to data.thresh to be consistent with other functions. Changes in openair 0.4-16 [2011-04-13] ===================================== * Fixed bugs in cutData when a partial year was used with type = "month". Also affected type = "season" and "weekday". The plots were labelled in the wrong order. * Ensure that missing wind sectors are skipped in type = "wd" for smoothTrend, MannKendall, timePlot and scatterPlot Changes in openair 0.4-15 [2011-04-12] ===================================== * Preparation for CRAN release * removed some restrictions on type/period combinations in polarAnnulus Changes in openair 0.4-14 [2011-04-05] ===================================== * Added preliminary greyscale method to openColours and linked in strip background and other text and line colour control for standard openair plots. Changes in openair 0.4-13 [2011-03-31] ===================================== * Removed ad-hoc sites from importKCL, added site classification to the help file and ensured GMT is exported. Changes in openair 0.4-12 [2011-03-22] ===================================== * Added y.relation option to smoothTrend and MannKendall to allow flexible y-scales. * NEW FUNCTION 'percentileRose' for flexible plotting of percentiles by wind direction * NEW FUNCTION 'selectRunning' for selecting run lengths of a numeric variable above a certain threshold. * Modified plot layout in polarPlot to maximise plotting area * Initial fixes to functions affected by time zone settings * polarFreq now has consistent annotation cf. windRose and percentileRose; new option grid.line to control radial axes. * Fixed couple of minor auto.text bugs in summaryPlot and timePlot * Removed dependency of date for polar functions where time-based types are not required. Changes in openair 0.4-11 [2011-03-18] ===================================== * More consistent use of strips in default plot (remove them) * Add percentage units to windRose and pollutionRose on radial scale Changes in openair 0.4-10 [2011-03-07] ===================================== * Fixed calendarPlot main/quickText handling. * Fixed calendarPlot first day of month error. * Fixed summaryPlot site ordering when more than one site. Changes in openair 0.4-9 [2011-03-04] ===================================== * Added new options to scatterPlot (plot.type, lwd, and lty) making it possible to add points and/or lines Changes in openair 0.4-8 [2011-02-24] ===================================== * Added new cutData type option 'daylight', enabled using new function cutDaylight. Changes in openair 0.4-7 [2011-02-16] ===================================== * Added new function conditionalQuantile for model evaluation purposes. * Fixed importAURN to account for a change in web domain address at AEA.