I have a .nc file that contains data every 6 hours of precipitation for 1 full year, my interest is to calculate the daily precipitation and compare with observed data, for this I must make them coincide temporally. To achieve this, the precipitation should accumulate between 12 utc of one day and 12 utc of the next day. Does anyone have a suggestion on how to achieve this with CDO?
Thank you!
Well if the first slice covers 12-18 UTC, then essentially you want to average the timeseries 4 slices at a time, (right?) in which case you can use this
cdo timselmean,4 infile.nc outfile.nc
If the timeseries starts instead at 00, you may need to remove the first two timeslices before you start (cdo seltimestep)
Another method is a bit more of a fudge, in that you can shift the series by 12 hours, and then use the day mean function. This would have the advantage of working for any frequency of data (i.e. you don't hardwire the factor "4" based on the data frequency)
cdo daymean -shifttime,-12hours infile.nc outfile.nc
The answer Adrian Tompkins gives should work well. One additional point to note is that you can remove time steps in CDO. So, if your time starts at 0 UTC ands at 24 UTC, you do not want the first and last time step of Adrian's first answer, but you could modify it as follows:
cdo -timselmean,4 -delete,timestep=-1,-2,1,2 infile.nc outfile.nc
This will require a 2.x version of CDO.
Related
I have a netCDF file for temperature going back the last 22 thousand years at a decadal average (TraCE dataset). I want to calculate 100 or 1000 year averages.
I am really stuck, if anyone could help then that would be great. I am mostly using R, but if it is simple in cdo then I can try this too.
I don't have any code to show as I really don't know where to start. Most examples I have seen have been on daily or yearly data... not decadal
Your data is decadal averages, so it should be easy to do this in CDO. You want to calculate a rolling average which is averaged over every 10 time steps. For this runmean is your friend. Just do the following:
cdo runmean,10 infile.nc outfile.nc
You might need to subset time afterwards, depending on the exact output you want. It sounds like the time you have may be non-standard, but runmean should still be OK.
Robert's solution is useful if you want a smoothed output at the 100 or 1000 year timescale. Your original dataset has 2200 timesteps, and runmean,10 smooths this and produces an output with 2200-9=2191 timesteps, each of which is an average over a 100 year window either centered on the slice itself or lagged/lead, depending on the option used.
However, from your question, I think you are more likely to want an output where the first slice is the average over the first century, the second is for the second century and so on, that is, an output with 220 timeslices, each a century average (or 22 time-slices of 1000 year averages). In other words, you want a command analogous to daymean, monmean and yearmean, but as there is no command called centurymean, then you can instead resort to the more generic command timselmean and manually define your window length:
# Centurial average:
cdo timselmean,10 infile.nc outfile.nc
# Millennial Average:
cdo timselmean,100 infile.nc outfile.nc
I think this should still work despite the non-CF compliant time units you mention in the comment (but let us know if it doesn't)
This is my first question on stackoverflow, sorry if the question is poorly put.
I am currently developing a project where I predict how much a person drinks each day. I currently have data that looks like this:
The menge column represents how much water a person has actually drunk in 30 minutes (So first value represents amount from 8:00 till before 8:30 etc..). This is a 1 day sample from 3 months of data. The day starts at 8 AM and ends at 8 PM.
I am trying to forecast the Time Series for each day. For example, given the first one or two time steps, we would predict the whole day and then we know how much in total the person has drunk until 8 PM.
I am trying to model this data as a Time Series object in R (Google Colab), in order to use Croston's Method for the forecasting. Using the ts() function, what should I set the frequency to knowing that:
The data is half-hourly
The data is from 8:00 till 20:00 each day (Does not span the whole day)
Would I need to make the data span the whole day by adding 0 values? Are there maybe better approaches for this? Thank you in advance.
When using the ts() function, the frequency is used to define the number of (usually regularly spaced) observations within a given time period. For your example, your observations are every 30 minutes between 8AM and 8PM, and your time period is 1 day. The time period of 1 day assumes that the patterns over each day is of most interest here, you could also use 1 week here.
So within each day of your data (8AM-8PM) you have 24 observations (24 half hours). So a suitable frequency for this data would be 24.
You can also pad the data with 0 values, however this isn't necessary and would complicate the model. If you padded the data so that it has observations for all half-hours of the day, the frequency would then be 48.
Before I resort to using Python, I would like to know if there is a simple way from the command line to convert an annual netCDF file into a file with daily data simply by duplication (including leap years), i.e. each annual value is duplicated 365 (366) times with the appropriate date stamp.
In the file each data value has the date stamped for the first day of each year:
cdo showdate population_ssp2soc_0p5deg_annual_2006-2100_density.nc4
2006-01-01 2007-01-01 2008-01-01 2009-01-01 2010-01-01 ...etc
I know it seems like a strange thing to do (the file size will be 365.25 times bigger!), but I want to read the data into a Fortran program that uses a daily timestep and do not want to mess around with dates in the Fortran code.
There might be a more efficient way of doing this. But you could first merge the original file and then a second which is the first, but time shifted to the final day of the year. Then temporally interpolate:
cdo -mergetime population_ssp2soc_0p5deg_annual_2006-2100_density.nc4 -shifttime,-1day -shifttime,1year population_ssp2soc_0p5deg_annual_2006-2100_density.nc4 temp.nc
cdo inttime,2006-01-01,12:00:00,1day temp.nc outfile.nc
I have climate data with a daily temporal resolution and would like a count of days that have precipitation (e.g., greater than 1mm/day) by month and by year.
I've tried eca_pd,1 and eca_rr1, but these commands return wet-day totals for all years.
For example, cdo eca_pd,1 infile outfile
Is there a command to return wet-days for each month and/or year?
You can accomplish this task with CDO's masking function, for more details beyond the answer below, you can also refer to my video guide on masking using cdo.
The first step is to make an equivalent file with 1 if P>threshold (1mm/day in your case) and 0 otherwise. For this we use the "greater than or equal to a constant" gec function (or ge="greater than" if you prefer):
cdo gec,1 input.nc mask.nc
(assuming units are mm/day in your input file).
Then you can simply sum this mask over the period (months, years etc) that you want your statistic
cdo monsum mask.nc nwetdays_mon.nc
cdo yearsum mask.nc nwetdays_year.nc
Of course you can pipe this if you like to do this on one line: e.g.
cdo monsum -gec,1 input.nc nwetdays_mon.nc
We can take this even further if you want to work out the climatology for a particular month. If you have a multiyear dataset then you can use the wonderful "ymonstat" commands. So for example, once you have calculated your monthly series of wet days above, you can calculate the average for each month with
cdo ymonmean nwetdays_mon.nc nwetdays_mon_clim.nc
You can then difference the series from this monthly climatology to give you the anomaly of wet days in each month over the series
cdo ymonsub nwetdays_mon.nc nwetdays_mon_clim.nc nwetdays_mon_anom.nc
I hope that helps!
(ps: I usually always find it is easier to calculate these kinds of statistics directly with CDO in this way, I rarely find that the built in climate functions calculate exactly the statistic as/how I want).
With NCO's ncap2, create a binary flag then total it in the desired dimension(s):
ncap2 -s 'rainy=(precip > 1);rainy_days=rainy.total($time)' in.nc out.nc
You can also do this in cf-python, essentially using the same methodology as the CDO example above, but in a Python environment, using the where and collapse methods:
import cf
# Read the dataset
f = cf.read('filename.nc')[0]
# Mask out dry days (assuming that your data
# units are 'mm day-1' or 'kg m-2 day-1', etc.)
wet = f.where(cf.le(1), cf.masked)
# If the data are in units of 'metres/day', say, then you could do:
# wet = f.where(cf.le(0.001), cf.masked)
# or
# wet = f.where(cf.le(1, 'mm day-1'), cf.masked)
# etc.
# Count the wet day occurrences by month
count_monthly = wet.collapse('T: sample_size', group=cf.M())
# Count the wet day occurrences by year
count_yearly = wet.collapse('T: sample_size', group=cf.Y())
# Get the data as numpy arrays
print(count_monthly.array)
print(count_yearly.array)
# Count the wet day totals by month
wet_day_sum_monthly = wet.collapse('T: sum', group=cf.M())
# Count the wet day totals by year
wet_day_sum_yearly = wet.collapse('T: sum', group=cf.Y())
What is the correct way to deal with datetimes in ggplot ?
I have data at several different dates and I would like to facet each date by the same time of day, e.g. between 1:30PM and 1:35PM, and plot the points between this time frame, how can I achieve this?
My data looks like:
datetime col1
2015-01-02 00:00:01 20
... ...
2015-01-02 11:59:59 34
2015-02-19 00:00:03 12
... ...
2015-02-19 11:59:58 27
I find myself often wanting to ggplot time series using datetime objects as the x-axis but I don't know how to use times only when dates aren't of interest.
The lubridate package will do the trick. There are commands you could use, specifically floor_date or ceiling_date to transform your datetime array.
I always use the chron package for times. It completely disregards dates and stores your time numerically (e.g. 1:30PM is stored as 13.5 because it's 13.5 hours into the day). That allows you to perform math on times, which is great for a lot of reasons, including calculating average time, the time between two points, etc.
For specific help with your plot you'll need to share a sample data frame in an easily copy-able format, and show the code you've tried so far.
This is a question I'd asked previously regarding the chron package, and it also gives an idea of how to share your data/ask a question that's easier for folks to reproduce and therefore answer:
Clear labeling of times class data on horizontal barplot/geom_segment