Skip to main content

This repo support auto line plot for multi-seed event file from TensorBoard

Project description

AutoLinePlotter

This repo support auto line plot for multi-seed event files from TensorBoard

To use this AutoLinePlotter, simply pip install AutoLinePlotter. Then, import plotter function from AutoLinePlotter to make a lineplot/lineplots.

For the function plotter, it has the following inputs:

root: str,: the location of the folder containing all algorithms data. Each is a folder containing a few seeds of event file generated from TensorBoard.

titles: List[str]: the titles to be plotted in each diagram; This has no effect in plot_together mode

labels: List[str]: the labels for each algorithm

x_axes: List[str]: the x-axis for each diagram; This has no effect in plot_together mode

y_axes: List[str]: the y-axis for each diagram; This has no effect in plot_together mode

plot_together: bool = False: whether to plot together or not

plot_together_x_axis: str = None: if plot_together, indicates the x axis for the plot

plot_together_y_axis: str = None: if plot_together, indicates the y axis for the plot

plot_together_title: str = None: if plot_together, indicates the title for the plot


For example, under User/files/pong directory (see below), we have two folders: pong_c51 and pong_dqn in which a few events files are included respectively. In this case, root should be 'User/files/pong'; titles could be ['pong_c51', 'pong_dqn']; labels could be ['pong_c51', 'pong_dqn']; x_axes could be ['env_step','env_step'] , y_axes could be ['return','return'].

  • User/files/pong

    • pong_c51
      • events.out.tfevents.1632662517.pong-c51-seed0
      • events.out.tfevents.1632662517.pong-c51-seed1
      • events.out.tfevents.1632662517.pong-c51-seed2
      • events.out.tfevents.1632662517.pong-c51-seed3
      • events.out.tfevents.1632662517.pong-c51-seed4
    • pong_dqn
      • events.out.tfevents.1632662517.pong-dqn-seed0
      • events.out.tfevents.1632662517.pong-dqn-seed1
      • events.out.tfevents.1632662517.pong-dqn-seed2
      • events.out.tfevents.1632662517.pong-dqn-seed3
      • events.out.tfevents.1632662517.pong-dqn-seed4

Funcion output

avatar avatar

Notice

titles, labels, x_axes, y_axes should follow the alphabetical order for the files within root. This is, in the case above, all these four list will be of dimension two. For whatever one enters, the first element in all these four lists should point to the final graph of pong_c51 and the second element in all these four lists should point to the final graph of pong_dqn.

The only restriction in naming folders is for the folders containing events files. The format should be env_algo. For instance, in the above case, the two folders are named pong_c51 and pong_dqn respectively. (there should not be any _ within env or algo naming)

If plot_together is enabled, these line plots will be plotted together. So, besides the above input, users should turn on plot_together and set plot_together_x_axis , plot_together_y_axis and plot_together_title. Notice that in this case, titles, x_axes, y_axes will have no effects. However, They should be implemented as if the user wants to plot with plot_together=False to avoid errors. If plot_together is enabled, the output for the above example is as follows:

Funcion output

avatar

Project details


Download files

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

Source Distribution

AutoLinePlotter-0.0.2.tar.gz (9.2 kB view details)

Uploaded Source

Built Distributions

AutoLinePlotter-0.0.2-py3.8.egg (9.7 kB view details)

Uploaded Source

AutoLinePlotter-0.0.2-py3-none-any.whl (9.7 kB view details)

Uploaded Python 3

File details

Details for the file AutoLinePlotter-0.0.2.tar.gz.

File metadata

  • Download URL: AutoLinePlotter-0.0.2.tar.gz
  • Upload date:
  • Size: 9.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.1 CPython/3.8.0

File hashes

Hashes for AutoLinePlotter-0.0.2.tar.gz
Algorithm Hash digest
SHA256 39569aa22a43d85621d6fe1f6f45dfa02a8151b3b760cc208991b1d94a14e458
MD5 18a03186cb684676731ee1d6a577651b
BLAKE2b-256 c5a57d9a01a403d76c95a4e252c0f8605f77b69e36edb828ded8eb22af2cce91

See more details on using hashes here.

File details

Details for the file AutoLinePlotter-0.0.2-py3.8.egg.

File metadata

  • Download URL: AutoLinePlotter-0.0.2-py3.8.egg
  • Upload date:
  • Size: 9.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.1 CPython/3.8.0

File hashes

Hashes for AutoLinePlotter-0.0.2-py3.8.egg
Algorithm Hash digest
SHA256 fc140343eae7ddab2055473c6803570798fc00dd0952048bf69fa24b1d3be3f2
MD5 53e91f6ca9c7622daf95f981186dcb25
BLAKE2b-256 e33df5818ebaf9490c24453277e3724915f12d31a97cfae6a839a1d7a213cb8c

See more details on using hashes here.

File details

Details for the file AutoLinePlotter-0.0.2-py3-none-any.whl.

File metadata

File hashes

Hashes for AutoLinePlotter-0.0.2-py3-none-any.whl
Algorithm Hash digest
SHA256 923af55c15ce33ab1c18442efd6464f7984ea82f363c7be7c70ba7065c12a9d5
MD5 0c64bb635968af32fd9ff40687c9353f
BLAKE2b-256 42599c82699975c33762a49cf1a50425b35c3fba4948663444f3d219c9896a5b

See more details on using hashes here.

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