ds format

Documentation version: 2.0.1 3.0.0 3.1.0 3.2.0 3.3.3 3.4.0 3.5.1 (latest) master
You are viewing documentation for version master. The latest release is 3.5.1.
GitHub

Command line

The ds-format Python package provides a command ds for reading, writing and modifying data files.

The command should be run in a shell such as Bash. zsh (the default shell on macOS) is not well supported due to the fact that curly brackets ({, }) are interpreted as special characters in this shell. zsh can be used if curly brackets are escaped with a backslash character (\).

On Unix-like operating systems, manual pages are available for the commands with man ds and man ds cmd. Note that you might have to add the manual page path (usually $HOME/.local/share/man/) to the MANPATH environment variable in order to access the manual pages.

Synopsis

ds

Tool for reading, writing and modifying dataset files.

Usage:

ds [cmd [args]] [options]
ds --help [cmd]
ds --version

The command line interface is based on the PST format. In all commands, variable, dimension and attribute names are interpreted as glob patterns, unless the -F option is enabled. Note that the pattern has to be enclosed in quotes in order to prevent the shell from interpreting the glob.

Arguments:

Options:

Available commands:

Supported input formats:

Supported output formats:

Environment variables:

Commands

Command Description
attrs Print attributes in a dataset.
cat Print variable data.
dims Print dimensions of a dataset or a variable.
ls List variables.
merge Merge files along a dimension.
meta Print dataset metadata.
rename Rename variables and attributes.
rename_dim Rename a dimension.
rm Remove variables or attributes.
select Select and subset variables.
set Set variable data, dimensions and attributes in an existing or new dataset.
size Print a variable size.
stats Print variable statistics.
type Print a variable type.

attrs

Print attributes in a dataset.

Usage: ds attrs [var] [attr] input [options]

The output is formatted as PST.

Arguments:

Examples:

Print dataset attributes in dataset.nc.

$ ds attrs dataset.nc
title: "Temperature data"

Print attributes of a variable temperature in dataset.nc.

$ ds attrs temperature dataset.nc
long_name: temperature units: celsius

Print a dataset attribute title.

$ ds attrs none title dataset.nc
"Temperature data"

Print an attribute units of a variable temperature.

$ ds attrs temperature units dataset.nc
celsius

cat

Print variable data.

Usage:

ds cat var input [options]
ds cat varinput [options]

Data are printed by the first index, one item per line, formatted as PST-formatted. If multiple variables are selected, items at a given index from all variables are printed on the same line as an array. The first line is a header containing a list of variables. Missing values are printed as empty rows (if printing one single dimensional variable) or as none.

Arguments:

Options:

Examples:

Print temperature values in dataset.nc.

$ ds cat temperature dataset.nc
temperature
16.000000
18.000000
21.000000

Print time and temperature values in dataset.nc.

$ ds cat time temperature dataset.nc
time temperature
1 16.000000
2 18.000000
3 21.000000

dims

Print dimensions of a dataset or a variable.

Usage: ds dims [var] input [options]

Arguments:

Options:

Examples:

Print dimensions of a dataset.

$ ds dims dataset.nc
time

Print dimensions of the variable temperature.

$ ds dims temperature dataset.nc
time

ls

List variables.

Usage:

ds [var]… input [options]
ds ls [var]… input [options]

Lines in the output are formatted as PST.

Arguments:

Options:

Examples:

Print a list of variables in dataset.nc.

$ ds ls dataset.nc
temperature
time

Print a detailed list of variables in dataset.nc.

$ ds ls -l dataset.nc
time: 3
temperature float64 { time }
time int64 { time }

Print a list of variables with an attribute units.

$ ds ls dataset.nc a: units
temperature celsius
time s

Print a list of variables with attributes long_name and units.

$ ds ls dataset.nc a: { long_name units }
temperature temperature celsius
time time s

Print all variables matching a glob “temp*” in dataset.nc.

$ ds ls 'temp*' dataset.nc
temperature

merge

Merge datasets along a dimension.

Usage: ds merge dim inputoutput [options]

Merge datasets along a dimension dim. If the dimension is not defined in the dataset, merge along a new dimension dim. If new is none and dim is not new, variables without the dimension dim are set with the first occurrence of the variable. If new is not none and dim is not new, variables without the dimension dim are merged along a new dimension new. If variables is not none, only those variables are merged along a new dimension, and other variables are set to the first occurrence of the variable. Variables which are merged along a new dimension and are not present in all datasets have their subsets corresponding to the datasets where they are missing filled with missing values. Dataset and variable metadata are merged sequentially from all datasets, with metadata from later datasets overriding metadata from the former ones.

Arguments:

Options:

Examples:

Write example data to dataset1.nc.

$ ds set { time none time { 1 2 3 } long_name: time units: s } { temperature none time { 16. 18. 21. } long_name: temperature units: celsius } title: "Temperature data" none dataset1.nc

Write example data to dataset2.nc.

$ ds set { time none time { 4 5 6 } long_name: time units: s } { temperature none time { 23. 25. 28. } long_name: temperature units: celsius } title: "Temperature data" none dataset2.nc

Merge dataset1.nc and dataset2.nc and write the result to dataset.nc.

$ ds merge time dataset1.nc dataset2.nc dataset.nc

Print time and temperature variables in dataset.nc.

$ ds cat time temperature dataset.nc
time temperature
1 16.000000
2 18.000000
3 21.000000
4 23.000000
5 25.000000
6 28.000000

meta

Print dataset metadata.

Usage: ds meta [var] input [options]

The output is formatted as PST.

Arguments:

Examples:

Print metadata of dataset.nc.

$ ds meta dataset.nc
.: {{
	title: "Temperature data"
}}
time: {{
	long_name: time
	units: s
	.dims: { time }
	.size: { 3 }
}}
temperature: {{
	long_name: temperature
	units: celsius
	.dims: { time }
	.size: { 3 }
}}

rename

Rename variables and attributes.

Usage:

ds rename vars input output [options]
ds rename var attrs input output [options]
ds rename { var attrs }input output [options]

Arguments:

Examples:

Rename variables time to newtime and temperature to newtemperature in dataset.nc and save the output in output.nc.

$ ds rename time: newtime temperature: newtemperature dataset.nc output.nc

Rename a dataset attribute title to newtitle in dataset.nc and save the output in output.nc.

$ ds rename none title: newtitle dataset.nc output.nc

Rename an attribute units of a variable temperature to newunits in dataset.nc and save the output in output.nc.

$ ds rename temperature units: newunits dataset.nc output.nc

rename_dim

Rename a dimension.

Usage:

ds rename_dim dims input output [options]

Arguments:

Examples:

Rename dimension time to newtime in dataset.nc and save the output in output.nc.

$ ds -l dataset.nc
time: 3
temperature
time
$ ds rename_dim time: newtime dataset.nc output.nc
$ ds -l output.nc
newtime: 3
temperature
time

rm

Remove variables or attributes.

Usage:

ds rm var input output [options]
ds rm var attr input output [options]

Arguments:

Examples:

Remove a variable temperature in dataset.nc and save the output in output.nc.

$ ds rm temperature dataset.nc output.nc

Remove variables time and temperature in dataset.nc and save the output in output.nc.

$ ds rm { time temperature } dataset.nc output.nc

Remove a dataset attribute title in dataset.nc and save the output in output.nc.

$ ds rm none title dataset.nc output.nc

Remove an attribute units of a variable temperature in dataset.nc and save the output in output.nc.

$ ds rm temperature units dataset.nc output.nc

select

Select and subset variables.

Usage: ds select [var…] [sel] input output [options]

select can also be used to convert between different file formats (ds select input output).

Arguments:

Examples:

Write data to dataset.nc.

$ ds set { time none time { 1 2 3 } long_name: time units: s } { temperature none time { 16. 18. 21. } long_name: temperature units: celsius } title: "Temperature data" none dataset.nc

List variables in dataset.nc.

$ ds dataset.nc
temperature
time

Select variable temperature from dataset.nc and write to temperature.nc.

$ ds select temperature dataset.nc temperature.nc

List variables in temperature.nc.

$ ds temperature.nc
temperature

Subset by time index 0 and write to 0.nc.

$ ds select time: 0 dataset.nc 0.nc

Print variables time and temperature in 0.nc.

$ ds cat time temperature 0.nc
time temperature
1 16.000000

Convert dataset.nc to JSON.

$ ds select dataset.nc dataset.json
$ cat dataset.json
{"time": [1, 2, 3], "temperature": [16.0, 18.0, 21.0], ".": {".": {"title": "Temperature data"}, "time": {"long_name": "time", "units": "s", ".dims": ["time"], ".size": [3]}, "temperature": {"long_name": "temperature", "units": "celsius ".dims": ["time"], ".size": [3]}}}

set

Set variable data, dimensions and attributes in an existing or new dataset.

Usage:

ds set ds_attrs input output [options]
ds set var [type [dims [data]]] [attrs]… input output [options]
ds set { var [type [dims [data]]] [attrs]… }ds_attrs input output [options]

Arguments:

Examples:

Write variables time and temperature to dataset.nc.

$ ds set { time none time { 1 2 3 } long_name: time units: s } { temperature none time { 16. 18. 21. } long_name: temperature units: celsius } title: "Temperature data" none dataset.nc

Set data of a variable temperature to an array of 16.0, 18.0, 21.0 in dataset.nc and save the output in output.nc.

$ ds set temperature none none { 16. 18. 21. } dataset.nc output.nc

Set a dimension of a variable temperature to time, data to an array of 16.0, 18.0, 21.0, its attribute long_name to “temperature” and units to “celsius” in dataset.nc and save the output in output.nc.

$ ds set temperature none time { 16. 18. 21. } long_name: temperature units: celsius dataset.nc output.nc

Set multiple variables in dataset.nc and save the output in output.nc.

$ ds set { time none time { 1 2 3 } long_name: time units: s } { temperature none time { 16. 18. 21. } long_name: temperature units: celsius } title: "Temperature data" dataset.nc output.nc

Set a dataset attribute newtitle to New title in dataset.nc and save the output in output.nc.

$ ds set newtitle: "New title" dataset.nc output.nc

Set an attribute newunits of a variable temperature to K in dataset.nc and save the output in output.nc.

$ ds set temperature newunits: K dataset.nc output.nc

size

Print a variable size.

Usage: ds size var input [options]

Arguments:

Examples:

Print the size of a variable temperature in a dataset dataset.nc.

$ ds size temperature dataset.nc
3

stats

Print variable statistics.

Usage: ds stats var input [options]

The output is formatted as PST.

Arguments:

Output description:

Examples:

Print statistics of variable temperature in dataset.nc.

$ ds stats temperature dataset.nc
count: 3 min: 16.000000 max: 21.000000 mean: 18.333333 median: 18.000000

type

Print a variable type.

Usage: ds type var input [options]

Arguments:

Examples:

Print the type of a variable temperature in a dataset dataset.nc.

$ ds type temperature dataset.nc
float64