similar to: [R-SIG-Finance] Removing outliers in tick data in R?

Sender: r-help-bounces at r-project.org On-Behalf-Of: comtech.usa at gmail.com Subject: Re: [R] [R-SIG-Finance] forecasting earnings, sales and gross margin of a company... Message-Id: <b1f16d9d0810231239k506d582i7ecb908b84bc1642 at mail.gmail.com> Recipient: ngottlieb at marinercapital.com -------------------------------------------------------- This information is being sent at the

I have a 10-column XLS file, with 2 date fields. As far as I can tell, they were configured identically in Excel 2010. One of these fields resembles "39406.577662037", whilst in Excel, it is shown as "2007-11-20 13:42:20". Applying as.Date() with the default format doesn't do it. Any ideas as to what format this is? Many thanks! -- H -- Sent from my mobile device Envoyait

I have a bunch of clean timeseries data obtained from a sensor and I'd like to apply a Kalman Filter to it to smoothe it out. Through a few days of Googling, reading papers, implementing such a filter in various languages, I finally realised that it may be built into R. So I did a "??kalman" at the R prompt and found that it is indeed there. However, the help page is a tad bare,

A sample of the data I have is: > head(sensor) logged_on accx accy accz compassx compassy compassz gyrox gyroy gyroz 1 1326561428000 -0.4602 0.8346 0.0936 0.145508 -0.350586 0.259766 59.617390 28.521740 59.617390 2 1326561428050 -0.4212 1.0452 0.1326 0.219727 -0.321289 0.241211 88.695656 27.478260 88.695656 3 1326561428100 -0.2496 1.3416 0.2886 0.214844 -0.326172

I don't know if timeboxing is the correct term to use to accomplish what I'm attempting, so allow me to explain. I have a set n of tagged observations in time series t. What I'm interested in is taking i seconds before and after every n. My code is below: # observations.xts is an xts time series and arg is the number of seconds to for the timebox timeboxes <-

My data: > dput(mydata) structure(list(V1 = c(1328565067, 1328565067.05, 1328565067.1, 1328565067.15, 1328565067.2, 1328565067.25), V2 = c(0.0963890795246276, 0.227296347215609, 0.240972698811569, 0.221208948983498, 0.230898231782485, 0.203282153087549), V3 = c(0.0245045248243853, 0.0835679411703579, 0.0612613120609633, 0.058568910563872, 0.0511868450318788, 0.0557714205674231 )), .Names =

I'd like to get the sum of every other row in a data.frame. When I actually set about doing this, I get the error in the subject line of this message. A sample of my data is below, followed by the function call that should give me the results I want: > dput(head(sens2)) structure(list(Time = c(1328565067, 1328565067.05, 1328565067.1, 1328565067.15, 1328565067.2, 1328565067.25), Y =

I have approximately 2.5 million rows from a number of sensor readings. Having plotted these, I can see a given pattern (say a spike in the amplitude away from the mean). I would now like to automate this procedure as we're expecting a great deal more data in the near future. Is there any package or function that will make this possible? Many thanks! I suppose, I could do something like:

I have a list of numbers corresponding to timestamps, a sample of which follows: c(1327211358, 1327221999, 1327527296, 1327555433, 1327701042, 1327761389, 1327780993, 1327815670, 1327822964, 1327897497, 1327897527, 1327937072, 1327938300, 1327957589, 1328044466, 1328127921, 1328157588, 1328213951, 1328236836, 1328300276, 1328335936, 1328429102) I would like to group these into hours. In other

> dput(sensor.sample) structure(c(1328565718.65, 1328566608.9, 1328566162.65, 1328566571.1, 1328566598.85, 1328565634.3, 1328566513.95, 1328565123.65, 1328565827.1, 1328566719.9, 1328565527.55, 1328565118.05, 1328565556.85, 1328565623.85, 1328565230.75, 1328566083.85, 1328566012.45, 1328566795.75, 1328565262.85, 1328566191.35, 1328565827.8, 1328566384.25, 1328565376.95, 1328566006.8,

I have two xts objects, call them "a" and "b", and am trying to merge them... > class(a) [1] "xts" "zoo" > class(b) [1] "xts" "zoo" > head(a) 2010-04-01 7.6343 2010-04-02 7.6343 2010-04-03 7.5458 2010-04-04 7.4532 2010-04-05 7.4040 2010-04-06 7.3317 > head(b) 2010-04-01 568.80 2010-04-05 571.01 2010-04-06

I have a list of bounds for a series of polygons. I do understand the formula to determine whether point i is within polygon X (X[x1] < i[x] & X[x2] > i[x] & X[y1] < i[y] & X[y2] > i[y]), and I can apply this throughout the dataset. However, this naive algorithm doesn't scale very well. The data set contains 10,000 points consisting of (n,e) pairs where I'm

Not sure if the question is appropos, but I have multiple csv's which are read into an xts object, corresponding to telemetry data (accelerometer, magnetometer/compass, and gyroscope). For examination, it would be ?ber useful if plot.zoo (or something similar) allowed me to zoom in and out of a subset of the time axes. -- Sent from my mobile device Envoyait de mon portable

Happy holidays all! I know it's very subjective to determine whether some data is outlier or not... But are there reasonally good and realistic methods of identifying outliers in R? Thanks a lot! [[alternative HTML version deleted]]

When I run the code below on Macintosh and Windows, the plot comes out fine. However, on Linux, the png generated is invalid from R console, and loading strucchange crashes rkward. Is this a known issue on Linux and, if so, is there a workaround? Many thanks! require(strucchange) data("RealInt") bp.ri <- breakpoints(RealInt~1, h=15) summary(bp.ri) fac.ri <- breakfactor(bp.ri,

I'm trying to read in a CSV, with lines looking like: HEADER, Latitude DecDeg, Latitude Hemisphere, Longitude DecDeg, Longitude Hemisphere, Speed knots, Bearing Degrees, fixQualityGga, noOfSatGga, altGga, heightGga, selectionGsa, fixGsa, pdopGsa, hdopGsa, vdopGsa, noOfSatGsv, Time, *checkSum $GPS, 3747.0224, N, 12223.4522, W, 0.36, 348.21, 1, 3, 4.01, 175.5, -25.2, A, 2, 4.14,

I was on vacation the last week and wrote some code to run a 500-day correlation between the Nasdaq tracking stock (QQQ) and 191 currency pairs for 500 days. The initial run took 9 hours(!) and I'd like to make it faster. So, I'm including my code below, in hopes that somebody will be able to figure out how to make it faster, either through parallelisation, or by making changes. I've

Hi all, I found that the PCA gave chaotic results when there are big changes in a few data points. Are there "improved" versions of PCA in R that can help with this problem? Please give me some pointers... Thank you! [[alternative HTML version deleted]]

Dear all, I have a time serie dataset such as the following with data acquired every 15 minutes: Date Heure Profondeur Temp?rature Salinit? Turbidit? Chloration 1 2012-07-06 08:47:22 -0.144 22.469 0.011 0.000 0 2 2012-07-06 09:02:21 -0.147 22.476 0.011 0.000 0 3 2012-07-06 09:17:21 -0.139 22.498 0.011 19.323 0 4 2012-07-06

Can't installe packag maDB or limma. Error is shown as Using R version 2.13.1, biocinstall version 2.8.4. Installing Bioconductor version 2.8 packages: Is there any other way to draw volcano plot ? Thanks [[alternative HTML version deleted]]