Skip to main content

Web Client for Visualizing Pandas Objects

Project description

image0

Live Demo


CircleCI PyPI ReadTheDocs codecov Downloads

What is it?

D-Tale is the combination of a Flask back-end and a React front-end to bring you an easy way to view & analyze Pandas data structures. It integrates seamlessly with ipython notebooks & python/ipython terminals. Currently this tool supports such Pandas objects as DataFrame, Series, MultiIndex, DatetimeIndex & RangeIndex.

Origins

D-Tale was the product of a SAS to Python conversion. What was originally a perl script wrapper on top of SAS’s insight function is now a lightweight web client on top of Pandas data structures.

In The News

Contents

Getting Started

PyCharm

jupyter

image6

image7

Installing the egg

# install dtale egg (important to use the "--upgrade" every time you install so it will grab the latest version)
$ pip install --upgrade dtale

Now you will have the ability to use D-Tale from the command-line or within a python-enabled terminal

Python Terminal

This comes courtesy of PyCharm image8 Feel free to invoke python or ipython directly and use the commands in the screenshot above and it should work

Issues With Windows Firewall

If you run into issues with viewing D-Tale in your browser on Windows please try making Python public under “Allowed Apps” in your Firewall configuration. Here is a nice article: How to Allow Apps to Communicate Through the Windows Firewall

Additional functions available programatically

import dtale
import pandas as pd

df = pd.DataFrame([dict(a=1,b=2,c=3)])

# Assigning a reference to a running D-Tale process
d = dtale.show(df)

# Accessing data associated with D-Tale process
tmp = d.data.copy()
tmp['d'] = 4

# Altering data associated with D-Tale process
# FYI: this will clear any front-end settings you have at the time for this process (filter, sorts, formatting)
d.data = tmp

# Shutting down D-Tale process
d.kill()

# using Python's `webbrowser` package it will try and open your server's default browser to this process
d.open_browser()

# There is also some helpful metadata about the process
d._data_id  # the process's data identifier
d._url  # the url to access the process

d2 = dtale.get_instance(d._data_id)  # returns a new reference to the instance running at that data_id

dtale.instances()  # prints a list of all ids & urls of running D-Tale sessions

Duplicate data check

To help guard against users loading the same data to D-Tale multiple times and thus eating up precious memory, we have a loose check for duplicate input data. The check runs the following: * Are row & column count the same as a previously loaded piece of data? * Are the names and order of columns the same as a previously loaded piece of data?

If both these conditions are true then you will be presented with an error and a link to the previously loaded data. Here is an example of how the interaction looks: image9

Jupyter Notebook

Within any jupyter (ipython) notebook executing a cell like this will display a small instance of D-Tale in the output cell. Here are some examples:

dtale.show

assignment

instance

image10

image11

image12

If you are running ipython<=5.0 then you also have the ability to adjust the size of your output cell for the most recent instance displayed:

image13

One thing of note is that a lot of the modal popups you see in the standard browser version will now open separate browser windows for spacial convienence:

Column Menus

Correlations

Describe

Histogram

Instances

image14

image15

image16

image17

image18

Google Colab & Kaggle

These are hosted notebook sites and thanks to the work of flask_ngrok users can run D-Tale within their notebooks.

DISCLAIMER: It is import that you set USE_NGROK to true when using D-Tale within these two services. Here is an example:

import pandas as pd

import dtale
import dtale.app as dtale_app

dtale_app.USE_NGROK = True

dtale.show(pd.DataFrame([1,2,3]))

Here are some video tutorials of each:

Service

Tutorial

Addtl Notes

Google Colab

image19

Kaggle

image20

make sure you switch the “Internet” toggle to “On” under settings of your notebook so you can install the egg from pip

R with Reticulate

I was able to get D-Tale running in R using reticulate. Here is an example:

library('reticulate')
dtale <- import('dtale')
df <- read.csv('https://vincentarelbundock.github.io/Rdatasets/csv/boot/acme.csv')
dtale$show(df, subprocess=FALSE, open_browser=TRUE)

Now the problem with doing this is that D-Tale is not running as a subprocess so it will block your R console and you’ll lose out the following functions: - manipulating the state of your data from your R console - adding more data to D-Tale

open_browser=TRUE isn’t required and won’t work if you don’t have a default browser installed on your machine. If you don’t use that parameter simply copy & paste the URL that gets printed to your console in the browser of your choice.

I’m going to do some more digging on why R doesn’t seem to like using python subprocesses (not sure if it something with how reticulate manages the state of python) and post any findings to this thread.

Here’s some helpful links for getting setup:

reticulate

installing python packages

Command-line

Base CLI options (run dtale --help to see all options available)

Prop

Description

--host

the name of the host you would like to use (most likely not needed since socket.gethostname() should figure this out)

--port

the port you would like to assign to your D-Tale instance

--name

an optional name you can assign to your D-Tale instance (this will be displayed in the <title> & Instances popup)

--debug

turn on Flask’s “debug” mode for your D-Tale instance

--no-reaper

flag to turn off auto-reaping subprocess (kill D-Tale instances after an hour of inactivity), good for long-running displays

--open-browser

flag to automatically open up your server’s default browser to your D-Tale instance

--force

flag to force D-Tale to try an kill any pre-existing process at the port you’ve specified so it can use it

Loading data from arctic(high performance datastore for pandas dataframes)

dtale --arctic-host mongodb://localhost:27027 --arctic-library jdoe.my_lib --arctic-node my_node --arctic-start 20130101 --arctic-end 20161231

Loading data from CSV

dtale --csv-path /home/jdoe/my_csv.csv --csv-parse_dates date

Loading data from JSON

dtale --json-path /home/jdoe/my_json.json --json-parse_dates date

or

dtale --json-path http://json-endpoint --json-parse_dates date

Loading data from a Custom loader - Using the DTALE_CLI_LOADERS environment variable, specify a path to a location containing some python modules - Any python module containing the global variables LOADER_KEY & LOADER_PROPS will be picked up as a custom loader - LOADER_KEY: the key that will be associated with your loader. By default you are given arctic & csv (if you use one of these are your key it will override these) - LOADER_PROPS: the individual props available to be specified. - For example, with arctic we have host, library, node, start & end. - If you leave this property as an empty list your loader will be treated as a flag. For example, instead of using all the arctic properties we would simply specify --arctic (this wouldn’t work well in arctic’s case since it depends on all those properties) - You will also need to specify a function with the following signature def find_loader(kwargs) which returns a function that returns a dataframe or None - Here is an example of a custom loader:

from dtale.cli.clickutils import get_loader_options

'''
  IMPORTANT!!! This global variable is required for building any customized CLI loader.
  When find loaders on startup it will search for any modules containing the global variable LOADER_KEY.
'''
LOADER_KEY = 'testdata'
LOADER_PROPS = ['rows', 'columns']


def test_data(rows, columns):
    import pandas as pd
    import numpy as np
    import random
    from past.utils import old_div
    from pandas.tseries.offsets import Day
    from dtale.utils import dict_merge
    import string

    now = pd.Timestamp(pd.Timestamp('now').date())
    dates = pd.date_range(now - Day(364), now)
    num_of_securities = max(old_div(rows, len(dates)), 1)  # always have at least one security
    securities = [
        dict(security_id=100000 + sec_id, int_val=random.randint(1, 100000000000),
             str_val=random.choice(string.ascii_letters) * 5)
        for sec_id in range(num_of_securities)
    ]
    data = pd.concat([
        pd.DataFrame([dict_merge(dict(date=date), sd) for sd in securities])
        for date in dates
    ], ignore_index=True)[['date', 'security_id', 'int_val', 'str_val']]

    col_names = ['Col{}'.format(c) for c in range(columns)]
    return pd.concat([data, pd.DataFrame(np.random.randn(len(data), columns), columns=col_names)], axis=1)


# IMPORTANT!!! This function is required for building any customized CLI loader.
def find_loader(kwargs):
    test_data_opts = get_loader_options(LOADER_KEY, kwargs)
    if len([f for f in test_data_opts.values() if f]):
        def _testdata_loader():
            return test_data(int(test_data_opts.get('rows', 1000500)), int(test_data_opts.get('columns', 96)))

        return _testdata_loader
    return None

In this example we simplying building a dataframe with some dummy data based on dimensions specified on the command-line: - --testdata-rows - --testdata-columns

Here’s how you would use this loader:

DTALE_CLI_LOADERS=./path_to_loaders bash -c 'dtale --testdata-rows 10 --testdata-columns 5'

UI

Once you have kicked off your D-Tale session please copy & paste the link on the last line of output in your browser image21

Dimensions/Main Menu

The information in the upper right-hand corner gives grid dimensions image22 - lower-left => row count - upper-right => column count - clicking the triangle displays the menu of standard functions (click outside menu to close it) image23

Column Menu Functions

image61

Moving Columns

image62

All column movements are saved on the server so refreshing your browser won’t lose them :ok_hand:

Hiding Columns

image63

All column movements are saved on the server so refreshing your browser won’t lose them :ok_hand:

Building Columns

image64

This video shows you how to build the following: - Numeric: adding/subtracting two columns or columns with static values - Bins: bucketing values using pandas cut & qcut as well as assigning custom labels - Dates: retrieving date properties (hour, weekday, month…) as well as conversions (month end)

Lock

Adds your column to “locked” columns - “locked” means that if you scroll horizontally these columns will stay pinned to the right-hand side - this is handy when you want to keep track of which date or security_id you’re looking at - by default, any index columns on the data passed to D-Tale will be locked

Unlock

Removed column from “locked” columns

Sorting

Applies/removes sorting (Ascending/Descending/Clear) to the column selected

Important: as you add sorts they sort added will be added to the end of the multi-sort. For example:

Action

Sort

click “a”

sort asc

a (asc)

click “b”

a (asc)

sort desc

a (asc), b(desc)

click “a”

a (asc), b(desc)

sort None

b(desc)

sort desc

b(desc), a(desc)

click “X” on sort display

Formats

Apply simple formats to numeric values in your grid

Type

Editing

Result

Numeric

image65

image66

Date

image67

image68

String

image69

image70

Here’s a grid of all the formats available with -123456.789 as input:

Format

Output

Precision (6)

-123456.789000

Thousands Sep

-123,456.789

Abbreviate

-123k

Exponent

-1e+5

BPS

-1234567890BPS

Red Negatives

-123457

Histogram

Display histograms in any number of bins (default: 20), simply type a new integer value in the bins input

image71

For Developers

Cloning

Clone the code (git clone ssh://git@github.com:manahl/dtale.git), then start the backend server:

$ git clone ssh://git@github.com:manahl/dtale.git
# install the dependencies
$ python setup.py develop
# start the server
$ python dtale --csv-path /home/jdoe/my_csv.csv --csv-parse_dates date

You can also run dtale from PyDev directly.

You will also want to import javascript dependencies and build the source:

$ npm install
# 1) a persistent server that serves the latest JS:
$ npm run watch
# 2) or one-off build:
$ npm run build

Running tests

The usual npm test command works:

$ npm test

You can run individual test files:

$ TEST=static/__tests__/dtale/DataViewer-base-test.jsx npm run test-file

Linting

You can lint all the JS and CSS to confirm there’s nothing obviously wrong with it:

$ npm run lint -s

You can also lint individual JS files:

$ npm run lint-js-file -s -- static/dtale/DataViewer.jsx

Formatting JS

You can auto-format code as follows:

$ npm run format

Docker Development

You can build python 27-3 & run D-Tale as follows:

$ yarn run build
$ docker-compose build dtale_2_7
$ docker run -it --network host dtale_2_7:latest
$ python
>>> import pandas as pd
>>> df = pd.DataFrame([dict(a=1,b=2,c=3)])
>>> import dtale
>>> dtale.show(df)

Then view your D-Tale instance in your browser using the link that gets printed

You can build python 36-1 & run D-Tale as follows:

$ yarn run build
$ docker-compose build dtale_3_6
$ docker run -it --network host dtale_3_6:latest
$ python
>>> import pandas as pd
>>> df = pd.DataFrame([dict(a=1,b=2,c=3)])
>>> import dtale
>>> dtale.show(df)

Then view your D-Tale instance in your browser using the link that gets printed

Startup Behavior

Here’s a little background on how the dtale.show() function works: - by default it will look for ports between 40000 & 49000, but you can change that range by specifying the environment variables DTALE_MIN_PORT & DTALE_MAX_PORT - think of sessions as python consoles or jupyter notebooks

  1. Session 1 executes dtale.show(df) our state is:

Session

Port

Active Data IDs

URL(s)

1

40000

1

http://localhost:40000/dtale/main/1

  1. Session 1 executes dtale.show(df) our state is:

Session

Port

Active Data IDs

URL(s)

1

40000

1,2

http://localhost:40000/dtale/main/[1,2]

  1. Session 2 executes dtale.show(df) our state is:

Session

Port

Active Data IDs

URL(s)

1

40000

1,2

http://localhost:40000/dtale/main/[1,2]

2

40001

1

http://localhost:40001/dtale/main/1

  1. Session 1 executes dtale.show(df, port=40001, force=True) our state is:

Session

Port

Active Data IDs

URL(s)

1

40001

1,2,3

http://localhost:40001/dtale/main/[1,2,3]

  1. Session 3 executes dtale.show(df) our state is:

Session

Port

Active Data IDs

URL(s)

1

40001

1,2,3

http://localhost:40001/dtale/main/[1,2,3]

3

40000

1

http://localhost:40000/dtale/main/1

  1. Session 2 executes dtale.show(df) our state is:

Session

Port

Active Data IDs

URL(s)

1

40001

1,2,3

http://localhost:40001/dtale/main/[1,2,3]

3

40000

1

http://localhost:40000/dtale/main/1

2

40002

1

http://localhost:40002/dtale/main/1

  1. Session 4 executes dtale.show(df, port=8080) our state is:

Session

Port

Active Data IDs

URL(s)

1

40001

1,2,3

http://localhost:40001/dtale/main/[1,2,3]

3

40000

1

http://localhost:40000/dtale/main/1

2

40002

1

http://localhost:40002/dtale/main/1

4

8080

1

http://localhost:8080/dtale/main/1

  1. Session 1 executes dtale.get_instance(1).kill() our state is:

Session

Port

Active Data IDs

URL(s)

1

40001

2,3

http://localhost:40001/dtale/main/[2,3]

3

40000

1

http://localhost:40000/dtale/main/1

2

40002

1

http://localhost:40002/dtale/main/1

4

8080

1

http://localhost:8080/dtale/main/1

  1. Session 5 sets DTALE_MIN_RANGE to 30000 and DTALE_MAX_RANGE 39000 and executes dtale.show(df) our state is:

Session

Port

Active Data ID(s)

URL(s)

1

40001

2,3

http://localhost:40001/dtale/main/[2,3]

3

40000

1

http://localhost:40000/dtale/main/1

2

40002

1

http://localhost:40002/dtale/main/1

4

8080

1

http://localhost:8080/dtale/main/1

5

30000

1

http://localhost:30000/dtale/main/1

Documentation

Have a look at the detailed documentation.

Requirements

D-Tale works with:

  • Back-end

    • arctic [extra]

    • dash

    • dash_daq

    • Flask

    • Flask-Compress

    • Pandas

    • scipy

    • six

  • Front-end

    • react-virtualized

    • chart.js

Acknowledgements

D-Tale has been under active development at Man Numeric since 2019.

Original concept and implementation: Andrew Schonfeld

Contributors:

Contributions welcome!

License

D-Tale is licensed under the GNU LGPL v2.1. A copy of which is included in LICENSE

Changelog

1.7.11 (2020-2-27)

  • hotfix for dash custom.js file missing from production webpack build script

1.7.10 (2020-2-27)

  • #75, added code snippet functionality to the following:

    • main grid, histogram, correlations, column building & charts

  • exposed CLI loaders through the following functions dtale.show_csv, dtale.show_json, dtale.show_arctic

    • build in such a way that it is easy for custom loaders to be exposed as well

  • #82, pinned future package to be >= 0.14.0

1.7.9 (2020-2-24)

  • support for google colab

  • bugfixes: #71, #72, #73

1.7.8 (2020-2-22)

  • #77, removal of multiprocessed timeouts

1.7.7 (2020-2-22)

  • centralized global state

1.7.6 (2020-2-21)

  • allowing the usage of context variables within filters

  • #64, handling for loading duplicate data to dtale.show

  • updated dtale.instances() to print urls rather than show all instances

  • removal of Dash “Export to png” function

  • passing data grid queries to chart page as default

  • added sys.exit() to the thread that manages the reaper

1.7.5 (2020-2-20)

  • hotfix for KeyError loading metadata for columns with min/max information

1.7.4 (2020-2-20)

  • #63: filtering columns with special characters in name

  • added json_loader CLI options

  • updated moving/locking of columns to be persisted to back-end as well as front-end

  • added the ability to show/hide columns

  • #61: added column builder popup

1.7.3 (2020-2-13)

  • added the ability to move columns left or right as well as to the front

  • added formatting capabilities for strings & dates

  • persist formatting settings to popup on reopening

  • bugfix for width-calculation on formatting change

1.7.2 (2020-2-12)

  • 60 timeout handling around chart requests

  • pre-loaded charts through URL search strings

  • pandas query examples in Filter popup

1.7.1 (2020-2-7)

  • added pie, 3D scatter & surface charts

  • updated popups to be displayed when the browser dimensions are too small to host a modal

  • removed Swagger due to its lack up support for updated dependencies

1.7.0 (2020-1-28)

  • redesign of charts popup to use plotly/dash

  • #55: raise exception when data contains duplicate column names

  • heatmap integration

  • combination of “_main.jsx” files into one for spacial optimization

  • #15: made arctic an “extra” dependency

1.6.10 (2020-1-12)

  • better front-end handling of dates for charting as to avoid timezone issues

  • the ability to switch between sorting any axis in bar charts

1.6.9 (2020-1-9)

  • bugfix for timezone issue around passing date filters to server for scatter charts in correlations popup

1.6.8 (2020-1-9)

  • additional information about how to use Correlations popup

  • handling of all-nan data in charts popup

  • styling issues on popups (especially Histogram)

  • removed auto-filtering on correlation popup

  • scatter point color change

  • added chart icon to cell that has been selected in correlation popup

  • responsiveness to scatter charts

  • handling of links to ‘main’,‘iframe’ & ‘popup’ missing data_id

  • handling of ‘inf’ values when getting min/max & describe data

  • added header to window popups (correlations, charts, …) and a link back to the grid

  • added egg building to cirleci script

  • correlation timeseries chart hover line

1.6.7 (2020-1-3)

  • #50: updates to rolling correlation functionality

1.6.6 (2020-1-2)

  • #47: selection of multiple columns for y-axis

  • updated histogram bin selection to be an input box for full customization

  • better display of timestamps in axis ticks for charts

  • sorting of bar charts by y-axis

  • #48: scatter charts in chart builder

  • “nunique” added to list of aggregations

  • turned on “threaded=True” for app.run to avoid hanging popups

  • #45: rolling computations as aggregations

  • Y-Axis editor

1.6.5 (2019-12-29)

  • test whether filters entered will return no data and block the user from apply those

  • allow for group values of type int or float to be displayed in charts popup

  • timeseries correlation values which return ‘nan’ will be replaced by zero for chart purposes

  • update ‘distribution’ to ‘series’ on charts so that missing dates will not show up as ticks

  • added “fork on github” flag for demo version & links to github/docs on “About” popup

  • limited lz4 to <= 2.2.1 in python 27-3 since latest version is no longer supported

1.6.4 (2019-12-26)

  • testing of hostname returned by socket.gethostname, use ‘localhost’ if it fails

  • removal of flask dev server banner when running in production environments

  • better handling of long strings in wordclouds

  • #43: only show timeseries correlations if datetime columns exist with multiple values per date

1.6.3 (2019-12-23)

  • updated versions of packages in yarn.lock due to issue with chart.js box & whisker plots

1.6.2 (2019-12-23)

  • #40: loading initial chart as non-line in chart builder

  • #41: double clicking cells in correlation grid for scatter will cause chart not to display

  • “Open Popup” button for ipython iframes

  • column width resizing on sorting

  • additional int/float descriptors (sum, median, mode, var, sem, skew, kurt)

  • wordcloud chart type

1.6.1 (2019-12-19)

  • bugfix for url display when running from command-line

1.6.0 (2019-12-19)

  • charts integration

    • the ability to look at data in line, bar, stacked bar & pie charts

    • the ability to group & aggregate data within the charts

  • direct ipython iframes to correlations & charts pages with pre-selected inputs

  • the ability to access instances from code by data id dtale.get_instance(data_id)

  • view all active data instances dtale.instances()

1.5.1 (2019-12-12)

  • conversion of new flask instance for each dtale.show call to serving all data associated with one parent process under the same flask instance unless otherwise specified by the user (the force parameter)

1.5.0 (2019-12-02)

  • ipython integration

    • ipython output cell adjustment

    • column-wise menu support

    • browser window popups for: Correlations, Coverage, Describe, Histogram & Instances

1.4.1 (2019-11-20)

  • #32: unpin jsonschema by moving flasgger to extras_require

1.4.0 (2019-11-19)

  • Correlations Pearson Matrix filters

  • “name” display in title tab

  • “Heat Map” toggle

  • dropped unused “Flask-Caching” requirement

1.3.7 (2019-11-12)

  • Bug fixes for:

    • #28: “Instances” menu option will now be displayed by default

    • #29: add hints to how users can navigate the correlations popup

    • add “unicode” as a string classification for column width calculation

1.3.6 (2019-11-08)

  • Bug fixes for:

    • choose between pandas.corr & numpy.corrcoef depending on presence of NaNs

    • hide timeseries correlations when date columns only contain one day

1.3.5 (2019-11-07)

  • Bug fixes for:

    • duplicate loading of histogram data

    • string serialization failing when mixing future.str & str in scatter function

1.3.4 (2019-11-07)

  • updated correlation calculation to use numpy.corrcoef for performance purposes

  • github rebranding from manahl -> man-group

1.3.3 (2019-11-05)

  • hotfix for failing test under certain versions of future package

1.3.2 (2019-11-05)

  • Bug fixes for:

    • display of histogram column information

    • reload of hidden “processes” input when loading instances data

    • correlations json failures on string conversion

1.3.1 (2019-10-29)

  • fix for incompatible str types when directly altering state of data in running D-Tale instance

1.3.0 (2019-10-29)

  • webbrowser integration (the ability to automatically open a webbrowser upon calling dtale.show())

  • flag for hiding the “Shutdown” button for long-running demos

  • “Instances” navigator popup for viewing all activate D-Tale instances for the current python process

1.2.0 (2019-10-24)

  • #20: fix for data being overriden with each new instance

  • #21: fix for displaying timestamps if they exist

  • calling show() now returns an object which can alter the state of a process

    • accessing/altering state through the data property

    • shutting down a process using the kill() function

1.1.1 (2019-10-23)

  • #13: fix for auto-detection of column widths for strings and floats

1.1.0 (2019-10-08)

  • IE support

  • Describe & About popups

  • Custom CLI support

1.0.0 (2019-09-06)

  • Initial public release

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

dtale-1.7.11.tar.gz (5.4 MB view hashes)

Uploaded Source

Built Distributions

dtale-1.7.11-py3.6.egg (5.5 MB view hashes)

Uploaded Source

dtale-1.7.11-py2.py3-none-any.whl (5.4 MB view hashes)

Uploaded Python 2 Python 3

dtale-1.7.11-py2.7.egg (5.5 MB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page