I give current day in a string in MATLAB. For example if today is '20180703', I need 7 strings containing:
'20180702'
'20180701'
'20180630'
'20180629'
'20180628'
'20180627'
'20180626'
Simple:
t = datetime('20180703', 'InputFormat', 'yyyyMMdd')
t = t - days(1:7)
datestr(t, 'yyyymmdd')
Edit.
As excaza pointed out, datetime and datestr use different input format. Hence, 'MM' in the first function, and 'mm' in the second one.
I would go with something like:
lastSevenDays = arrayfun(#(offset) datestr(now-offset, 'yyyymmdd'), 1:7, 'UniformOutput', false)
or more matlaby:
datestr(now - days(1:7), 'yyyymmdd')
Below is my avg_df:
Date Model INumber Type TimeDiff Device
326 20/07/18 TG I625 Devicetime 0:02:31 RD
328 20/07/18 TG I5271 Devicetime 0:00:32 RD
332 20/07/18 TG I660 Devicetime 0:00:31 RD
I want to get average of "TimeDiff". I know that i can convert Time into secs and get avg and can format it back, but would be interested to know if there is any way that i can get without formatting time back and forth. something like below:
print(avg_df.loc[:,"TimeDiff"].mean())
Appreciate any help!
You can get the average if you convert it to timedelta first:
>>> pd.to_timedelta(df['TimeDiff']).mean()
Timedelta('0 days 00:01:11.333333')
Trying to convert data at daily frequency to weekly frequency.
In:
weeklyaaapl = pd.DataFrame()
weeklyaapl['Open'] = aapl.Open.resample('W').iloc[0]
#here I am trying to take the first value of the aapl.Open,
#that falls within the week.
Out:
ValueError: .resample() is now a deferred operation
use .resample(...).mean() instead of .resample(...)
I want the true open (the first open that prints for the week) (the open of the first day in that week).
It instead wants me to take the mean of the daily open values for a given week using .mean(), which is not the information I need.
Can't seem to interpret the error, documentation isn't helping either.
I think you need.
aapl.resample('W').first()
Output:
Open High Low Close Volume
Date
2010-01-10 30.49 30.64 30.34 30.57 123432050
2010-01-17 30.40 30.43 29.78 30.02 115557365
2010-01-24 29.76 30.74 29.61 30.72 182501620
2010-01-31 28.93 29.24 28.60 29.01 266424802
2010-02-07 27.48 28.00 27.33 27.82 187468421
There must be a quick solution for this, but after 30min I gave up and need help.
This is the format of source data
0h56m40s 0h57m10s 1h00m40s 1h02m15s 1h02m25s
52m47s 54m25s 54m52s 57m23s 57m43s
49m30s 54m31s 54m34s 56m35s 56m36s
47m45s 48m03s 51m02s 52m23s 53m05s
46m54s 49m29s 50m51s 51m02s 51m03s
46m09s 47m56s 50m16s 51m20s 51m53s
46m55s 47m08s 47m13s 48m16s 50m11s
and I need this in time format like 0h56m40s to 0:56:40
I tried search/replace, from h to :, m to : and removing s, works for when there's hour, but messes up for when only minutes are there.
Any tips?
You can concatenate 0: if the input string is too short:
=(IF(LEN(A1)<7,"0:","") & SUBSTITUTE(SUBSTITUTE(SUBSTITUTE(A1,"h",":"),"m",":"),"s",""))+0
The +0 part converts string to time value (change cell format to h:mm:ss). If you prefer to keep it in text format, remove +0.
How would I extract the time from a series of POSIXct objects discarding the date part?
For instance, I have:
times <- structure(c(1331086009.50098, 1331091427.42461, 1331252565.99979,
1331252675.81601, 1331262597.72474, 1331262641.11786, 1331269557.4059,
1331278779.26727, 1331448476.96126, 1331452596.13806), class = c("POSIXct",
"POSIXt"))
which corresponds to these dates:
"2012-03-07 03:06:49 CET" "2012-03-07 04:37:07 CET"
"2012-03-09 01:22:45 CET" "2012-03-09 01:24:35 CET"
"2012-03-09 04:09:57 CET" "2012-03-09 04:10:41 CET"
"2012-03-09 06:05:57 CET" "2012-03-09 08:39:39 CET"
"2012-03-11 07:47:56 CET" "2012-03-11 08:56:36 CET"
Now, I have some values for a parameter measured at those times:
val <- c(1.25343125e-05, 0.00022890575,
3.9269125e-05, 0.0002285681875,
4.26353125e-05, 5.982625e-05,
2.09575e-05, 0.0001516951251,
2.653125e-05, 0.0001021391875)
I would like to plot val vs time of the day, irrespectively of the specific day when val was measured.
Is there a specific function that would allow me to do that?
You can use strftime to convert datetimes to any character format:
> t <- strftime(times, format="%H:%M:%S")
> t
[1] "02:06:49" "03:37:07" "00:22:45" "00:24:35" "03:09:57" "03:10:41"
[7] "05:05:57" "07:39:39" "06:47:56" "07:56:36"
But that doesn't help very much, since you want to plot your data. One workaround is to strip the date element from your times, and then to add an identical date to all of your times:
> xx <- as.POSIXct(t, format="%H:%M:%S")
> xx
[1] "2012-03-23 02:06:49 GMT" "2012-03-23 03:37:07 GMT"
[3] "2012-03-23 00:22:45 GMT" "2012-03-23 00:24:35 GMT"
[5] "2012-03-23 03:09:57 GMT" "2012-03-23 03:10:41 GMT"
[7] "2012-03-23 05:05:57 GMT" "2012-03-23 07:39:39 GMT"
[9] "2012-03-23 06:47:56 GMT" "2012-03-23 07:56:36 GMT"
Now you can use these datetime objects in your plot:
plot(xx, rnorm(length(xx)), xlab="Time", ylab="Random value")
For more help, see ?DateTimeClasses
The data.table package has a function 'as.ITime', which can do this efficiently use below:
library(data.table)
x <- "2012-03-07 03:06:49 CET"
as.IDate(x) # Output is "2012-03-07"
as.ITime(x) # Output is "03:06:49"
There have been previous answers that showed the trick. In essence:
you must retain POSIXct types to take advantage of all the existing plotting functions
if you want to 'overlay' several days worth on a single plot, highlighting the intra-daily variation, the best trick is too ...
impose the same day (and month and even year if need be, which is not the case here)
which you can do by overriding the day-of-month and month components when in POSIXlt representation, or just by offsetting the 'delta' relative to 0:00:00 between the different days.
So with times and val as helpfully provided by you:
## impose month and day based on first obs
ntimes <- as.POSIXlt(times) # convert to 'POSIX list type'
ntimes$mday <- ntimes[1]$mday # and $mon if it differs too
ntimes <- as.POSIXct(ntimes) # convert back
par(mfrow=c(2,1))
plot(times,val) # old times
plot(ntimes,val) # new times
yields this contrasting the original and modified time scales:
Here's an update for those looking for a tidyverse method to extract hh:mm::ss.sssss from a POSIXct object. Note that time zone is not included in the output.
library(hms)
as_hms(times)
Many solutions have been provided, but I have not seen this one, which uses package chron:
hours = times(strftime(times, format="%T"))
plot(val~hours)
(sorry, I am not entitled to post an image, you'll have to plot it yourself)
I can't find anything that deals with clock times exactly, so I'd just use some functions from package:lubridate and work with seconds-since-midnight:
require(lubridate)
clockS = function(t){hour(t)*3600+minute(t)*60+second(t)}
plot(clockS(times),val)
You might then want to look at some of the axis code to figure out how to label axes nicely.
The time_t value for midnight GMT is always divisible by 86400 (24 * 3600). The value for seconds-since-midnight GMT is thus time %% 86400.
The hour in GMT is (time %% 86400) / 3600 and this can be used as the x-axis of the plot:
plot((as.numeric(times) %% 86400)/3600, val)
To adjust for a time zone, adjust the time before taking the modulus, by adding the number of seconds that your time zone is ahead of GMT. For example, US central daylight saving time (CDT) is 5 hours behind GMT. To plot against the time in CDT, the following expression is used:
plot(((as.numeric(times) - 5*3600) %% 86400)/3600, val)