{ "id": 341898, "title": "🎨 How to make a Custom Colormap for matplotlib graphs", "url": "/competitions/rsna-2022-cervical-spine-fracture-detection/discussion/341898", "author_name": "Andrada", "post_date": "2022-08-04T17:13:30.261000", "votes": 21, "comment_count": 5, "views": 0, "content": "

Ok so I have been searching a bunch for a way to create a custom map and what is the easiest and quickest way to do so.

\n

There are many premade colormaps for matplotlib but I usually create my own categorical color scheme for each competition.

\n

Therefore, I wanted to use this personalized palette in my graphs too (below I've a picture with the palette I am using for this competition):
\n

\n

So, here's how I do this quickly.

\n

1. The initial graph

\n

By default, the below scatterplot looks like this:

\n
sns.scatterplot(data=data, x=\"count\", y=\"y\", size=\"count\", alpha=0.65, sizes=(500, 2000),\n               hue=\"count\")\n
\n

\n

2. Create Customized Map

\n

Next step is to create a CMAP1 that's personalized with my_colors:

\n
from matplotlib.colors import ListedColormap\n\nmy_colors = [\"#5EAFD9\", \"#449DD1\", \"#3977BB\", \n             \"#2D51A5\", \"#5C4C8F\", \"#8B4679\",\n             \"#C53D4C\", \"#E23836\", \"#FF4633\", \"#FF5746\"]\nCMAP1 = ListedColormap(my_colors)\n
\n

3. Change the final graph

\n
sns.scatterplot(data=data, x=\"count\", y=\"y\", size=\"count\", alpha=0.65, sizes=(500, 2000),\n               hue=\"count\", palette=CMAP1)\n
\n

\n

Hope this helps anyone interested 😊 Cheers!

", "messages": [ { "id": 1884831, "postDate": "2022-08-04T17:13:30.260Z", "content": "

Ok so I have been searching a bunch for a way to create a custom map and what is the easiest and quickest way to do so.

\n

There are many premade colormaps for matplotlib but I usually create my own categorical color scheme for each competition.

\n

Therefore, I wanted to use this personalized palette in my graphs too (below I've a picture with the palette I am using for this competition):
\n

\n

So, here's how I do this quickly.

\n

1. The initial graph

\n

By default, the below scatterplot looks like this:

\n
sns.scatterplot(data=data, x=\"count\", y=\"y\", size=\"count\", alpha=0.65, sizes=(500, 2000),\n               hue=\"count\")\n
\n

\n

2. Create Customized Map

\n

Next step is to create a CMAP1 that's personalized with my_colors:

\n
from matplotlib.colors import ListedColormap\n\nmy_colors = [\"#5EAFD9\", \"#449DD1\", \"#3977BB\", \n             \"#2D51A5\", \"#5C4C8F\", \"#8B4679\",\n             \"#C53D4C\", \"#E23836\", \"#FF4633\", \"#FF5746\"]\nCMAP1 = ListedColormap(my_colors)\n
\n

3. Change the final graph

\n
sns.scatterplot(data=data, x=\"count\", y=\"y\", size=\"count\", alpha=0.65, sizes=(500, 2000),\n               hue=\"count\", palette=CMAP1)\n
\n

\n

Hope this helps anyone interested 😊 Cheers!

", "rawMarkdown": "Ok so I have been searching a bunch for a way to create a custom map and what is the **easiest** and **quickest** way to do so.\n\nThere are many [premade colormaps for matplotlib](https://matplotlib.org/stable/tutorials/colors/colormaps.html) but I usually create my own *categorical* color scheme for each competition.\n\nTherefore, I wanted to use this personalized palette in my graphs too (below I've a picture with the palette I am using for this competition):\n\n\nSo, here's how I do this quickly.\n\n#### 1. The initial graph\n\nBy default, the below scatterplot looks like this:\n\n```\nsns.scatterplot(data=data, x=\"count\", y=\"y\", size=\"count\", alpha=0.65, sizes=(500, 2000),\n hue=\"count\")\n```\n\n\n#### 2. Create Customized Map\n\nNext step is to create a `CMAP1` that's personalized with `my_colors`:\n\n```\nfrom matplotlib.colors import ListedColormap\n\nmy_colors = [\"#5EAFD9\", \"#449DD1\", \"#3977BB\", \n \"#2D51A5\", \"#5C4C8F\", \"#8B4679\",\n \"#C53D4C\", \"#E23836\", \"#FF4633\", \"#FF5746\"]\nCMAP1 = ListedColormap(my_colors)\n```\n\n#### 3. Change the final graph\n\n```\nsns.scatterplot(data=data, x=\"count\", y=\"y\", size=\"count\", alpha=0.65, sizes=(500, 2000),\n hue=\"count\", palette=CMAP1)\n```\n\n\nHope this helps anyone interested 😊 Cheers!", "votes": 20 }, { "id": 1885281, "postDate": "2022-08-05T04:13:35.177Z", "content": "

very useful information ! Thank you for sharing !

", "rawMarkdown": "very useful information ! Thank you for sharing !" }, { "id": 1884893, "postDate": "2022-08-04T18:09:44.380Z", "content": "

This is quite useful to make one's work look great and perhaps elicit key features also! Great work!! Bookmarked!!

", "rawMarkdown": "This is quite useful to make one's work look great and perhaps elicit key features also! Great work!! Bookmarked!!" }, { "id": 1890978, "postDate": "2022-08-09T06:52:58.123Z", "content": "

You always do something unique @andradaolteanu

", "rawMarkdown": "You always do something unique @andradaolteanu " }, { "id": 2042201, "postDate": "2022-11-24T14:02:59.350Z", "content": "

This is gold! Thank you!

", "rawMarkdown": "This is gold! Thank you!" }, { "id": 1888833, "postDate": "2022-08-07T19:49:30.833Z", "content": "

that's helpful, thanks!

", "rawMarkdown": "that's helpful, thanks!" } ], "comments": [ { "id": 1885281, "author_name": "Arun Purakkatt", "author_url": "", "post_date": "2022-08-05T04:13:35.177000", "content": "

very useful information ! Thank you for sharing !

", "votes": 0, "replies": [] }, { "id": 1884893, "author_name": "Ravi Ramakrishnan", "author_url": "", "post_date": "2022-08-04T18:09:44.380000", "content": "

This is quite useful to make one's work look great and perhaps elicit key features also! Great work!! Bookmarked!!

", "votes": 0, "replies": [] }, { "id": 1890978, "author_name": "Abdullah Sangha", "author_url": "", "post_date": "2022-08-09T06:52:58.123000", "content": "

You always do something unique @andradaolteanu

", "votes": 0, "replies": [] }, { "id": 2042201, "author_name": "Leandro Destefani", "author_url": "", "post_date": "2022-11-24T14:02:59.350000", "content": "

This is gold! Thank you!

", "votes": 0, "replies": [] }, { "id": 1888833, "author_name": "Ahmed Yassin", "author_url": "", "post_date": "2022-08-07T19:49:30.833000", "content": "

that's helpful, thanks!

", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "1884831": "Ok so I have been searching a bunch for a way to create a custom map and what is the **easiest** and **quickest** way to do so.\n\nThere are many [premade colormaps for matplotlib](https://matplotlib.org/stable/tutorials/colors/colormaps.html) but I usually create my own *categorical* color scheme for each competition.\n\nTherefore, I wanted to use this personalized palette in my graphs too (below I've a picture with the palette I am using for this competition):\n\n\nSo, here's how I do this quickly.\n\n#### 1. The initial graph\n\nBy default, the below scatterplot looks like this:\n\n```\nsns.scatterplot(data=data, x=\"count\", y=\"y\", size=\"count\", alpha=0.65, sizes=(500, 2000),\n hue=\"count\")\n```\n\n\n#### 2. Create Customized Map\n\nNext step is to create a `CMAP1` that's personalized with `my_colors`:\n\n```\nfrom matplotlib.colors import ListedColormap\n\nmy_colors = [\"#5EAFD9\", \"#449DD1\", \"#3977BB\", \n \"#2D51A5\", \"#5C4C8F\", \"#8B4679\",\n \"#C53D4C\", \"#E23836\", \"#FF4633\", \"#FF5746\"]\nCMAP1 = ListedColormap(my_colors)\n```\n\n#### 3. Change the final graph\n\n```\nsns.scatterplot(data=data, x=\"count\", y=\"y\", size=\"count\", alpha=0.65, sizes=(500, 2000),\n hue=\"count\", palette=CMAP1)\n```\n\n\nHope this helps anyone interested 😊 Cheers!", "1885281": "very useful information ! Thank you for sharing !", "1884893": "This is quite useful to make one's work look great and perhaps elicit key features also! Great work!! Bookmarked!!", "1890978": "You always do something unique @andradaolteanu ", "2042201": "This is gold! Thank you!", "1888833": "that's helpful, thanks!" } }