csvtools

CLI tools useful to work with csv files. Below is the documentation for each tool.

csvpp

usage: csvpp [-h] [-q] [--careful] [-s SEPARATOR] [-n LINES_NUMBER] [-f] [-o OUTPUT_FILE] [file]

Print csv file in human-readable format. Input is taken from STDIN by default. The first line is header. It contains names of fields.

positional arguments: file

File to read input from. stdin is used by default

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-s SEPARATOR, --separator SEPARATOR

Separator to be used

-n LINES_NUMBER, --lines_number LINES_NUMBER

Number of lines used to set column width

-f, --format_floats

Format floating-point numbers nicely

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

examples: cat file.txt | csvpp -f | less -SR

csvcut

usage: csvcut [-h] [-q] [--careful] [-s SEPARATOR] [-o OUTPUT_FILE] [-f FIELDS] [-c] [-u] [file]

Select some columns from csv streem. Could change order of fields.

positional arguments: file

File to read input from. stdin is used by default

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-s SEPARATOR, --separator SEPARATOR

Separator to be used

-f FIELDS, --fields FIELDS

Specify list of fields (comma separated) to cut. Field names or field numbers can be used. Dash can be used to specify fields ranges. Range 'F1-F2' stands for all fields between F1 and F2. Range '-F2' stands for all fields up to F2. Range 'F1-' stands for all fields from F1 til the end.

-c, --complement

Instead of leaving only specified columns, leave all except specified.

-u,--unique

Remove duplicates from list of FIELDS

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

examples: csvcut -f 1,2 stat.txt csvcut -f st,shows,clicks stat.txt cat stat.txt | csvcut -f shows,uniq,clicks cat stat.txt | csvcut -f select_type-clicks all fields from select_type to clicks cat stats.txt | csvcut -f -shows stat.txt all fields from the first till shows csvcut -f page_id- all fields from page_id till the end csvcut -f description --complement all fields except for description

csvhead

usage: csvhead [-h] [-q] [--careful] [-o OUTPUT_FILE] [-n ROWS_COUNT] [file]

Print header and first lines of input.

positional arguments: file

File to read input from. stdin is used by default

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-n ROWS_COUNT, --number_of_lines ROWS_COUNT

Number of first rows to print

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

examples: cat file.csv | csvhead -n 100 prints first 100 rows of file.csv

csvtail

usage: csvtail [-h] [-q] [--careful] [-o OUTPUT_FILE] [-n ROWS_COUNT] [file]

Print header and last lines of input.

positional arguments: file

File to read input from. stdin is used by default

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-n ROWS_COUNT, --number_of_lines ROWS_COUNT

Number of last rows to print if positive ROWS_COUNT. Else skips ROWS_COUNT lines and prints till the end of input.

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

examples: cat file.csv | csvtail -n -100 skip first 100 rows and print file.csv till the end.

csvmap

usage: csvmap [-h] [-q] [--careful] [-s SEPARATOR] [-o OUTPUT_FILE] [-e EXEC] expression [file]

Transform each row of a csv file with an expression provided.

positional arguments: expression

Python expression to be used to transform a row. Specific columns can be refered as a fields of row object named r

file

File to read input from. stdin is used by default

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-s SEPARATOR, --separator SEPARATOR

Separator to be used

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

-e EXEC, --exec EXEC

Execute python code before starting the transformation. Might be useful for import statements or even for python functions definition.

examples: cat file.csv | csvmap 'r.ratio = r.a / r.b, r.b *= 1000' assuming that file.csv contain columns a and b, introduces a new column ratio containing the a / b. Also multiplies column b by 1000.

cat file.csv | csvmap 'r.root_a = sqrt(r.a)' --exec 'from numpy import sqrt' assuming that file.csv contains column a, introduces a new column root_a containing sqrt(a). Prior to running the row transformation, imports sqrt function from numpy.

csvreduce

usage: csvreduce [-h] [-q] [--careful] [-s SEPARATOR] [-o OUTPUT_FILE] [-k KEYS] [-a AGGREGATORS] [file]

Reduces csv file using the KEYS provided. Reducing is a process of aggregating rows with the same keys by applying AGGREGATORS to them. In other words, the rows will be grouped by the KEYS during the aggregation process.

positional arguments: file

File to read input from. stdin is used by default

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-s SEPARATOR, --separator SEPARATOR

Separator to be used

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

-k KEYS, --keys KEYS

Comma-separated list of columns to be used as reduce keys. Column names or column numbers can be used here

-a AGGREGATORS, --agregators AGGREGATORS

Comma-separated list of value-aggregators. Each aggregator might be one of the following: first, last, sum, mean, min, max, std (standard deviation), count. Each aggragator (except count) is a function expecting 2 arguments: column name or number and the resulting field name. The resulting field name has a default value of $AGGREGATOR_NAME_$FIRST_ARGUMENT' (e.g. for sum('a') it will have a default value of sum_a). Please see the examples for more details.

examples: cat csvle.txt | csvreduce -r sum('price', 'overall_price'),count assuming that csvle.txt has a column named price, the result will be a csvle containing just a single row with columns named overall_price and count.

`cat flat_prices.csv | csvreduce -k type,district -a max('price'),min('price'),mean('price'),avg('square')`

assuming that flat_prices.csv has columns name price and district, the result will be a csvle containing the maximum, minimum and average price and average square of a flat for each district and commercial type.

csvsort

usage: csvsort [-h] [-q] [--careful] [-s SEPARATOR] [-o OUTPUT_FILE] [-k KEYS] [-m MAX_ROWS] [--descending] [--numeric] [file]

Sort the rows of csv stream ascending.

positional arguments: file

File to read input from. stdin is used by default

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-s SEPARATOR, --separator SEPARATOR

Separator to be used

-k KEYS, --keys KEYS

Specify the list of keys (comma separated) to sort on. Field names or field numbers can be used. Dash can be used to specify fields ranges. Range 'F1-F2' stands for all fields between F1 and F2. Range '-F2' stands for all fields up to F2. Range 'F1-' stands for all fields from F1 til the end.

--descending

If provided, perform descending sort instead of ascending

--numeric

If provided, keys will be interpreted as numbers. Otherwise - as strings.

-m MAX_ROWS, --max-rows MAX_ROWS

Don't load to memory more than MAX_ROWS rows at a time. This option is crucial if you have to deal with huge csv files. Default value is 0 that meanse that this will sort file in memory.

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

examples: cat stat.csv | csvsort -k shows

csvplot

usage: csvplot [-h] [-q] [--careful] [-s SEPARATOR] [-o OUTPUT_FILE] [-x KEY] [-y KEYS] [--xlabel LABEL] [-ylabels LABELS] [file]

Plot the data based on csv file contents.

positional arguments: file

File to read input from. stdin is used by default

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-s SEPARATOR, --separator SEPARATOR

Separator to be used

-x KEY

Specify the key to iterate over x-axes. If not provided, use row number instead.

--xlabel LABEL

Label to be used for X axis.

-y KEYS

Specify columns to be plotted. One line graph will be plotted per each column. The plots will have different colors.

--ylabel LABELS

Labeles to be used for Y axis.

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

examples: cat stat.csv | csvsort -k shows

csvjoin

usage: csvjoin [-h] [-q] [--careful] [-s SEPARATOR] [-o OUTPUT_FILE] [-k KEYS] [-c CONFLICT_PREFIX] [-t TYPE] file2 [file]

Join two csv files using the keys provided.

positional arguments: file

File to read input from. stdin is used by default

file2

File to join with

optional arguments: -h, --help

show help message and exit

-q, --quiet

Don't print information regarding errors

--careful

Stop if input contains an incorrect row

-s SEPARATOR, --separator SEPARATOR

Separator to be used

-k KEYS, --fields KEYS

Comma-separated list of columns to be used as reduce keys. Column names or column numbers can be used here. These columns will be used as the join keys.

-c CONFLICT_PREFIX, --conflict_prefix CONFLICT_PREFIX

Specify a prefix to be used for the columns taken from file2 having the same names. Default is conflict_

-t TYPE, --type TYPE

Type of join to be used. Either INNER, LEFT or OUTER. INNER is default.

-o OUTPUT_FILE, --output_file OUTPUT_FILE

Output file. stdout is used by default

examples: csvjoin stat1.txt stat2.txt -k id, -t OUTER