{"cells":[{"metadata":{},"cell_type":"markdown","source":"> Idea from : [Bengali.AI Handwritten Grapheme - Getting Started](https://www.kaggle.com/gpreda/bengali-ai-handwritten-grapheme-getting-started)\n\nUsing heatmap, I thought about a visualization that could help someone who doesn't know this language.\n\nWouldn't it be easier to understand this data if you had an interactive tooltip?"},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"import pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt\n\nimport seaborn as sns\nfrom PIL import Image\n\nfrom bokeh.plotting import figure, show, output_notebook\nfrom bokeh.models import HoverTool, ColumnDataSource, LinearColorMapper, BasicTicker, PrintfTickFormatter, ColorBar\nfrom bokeh.palettes import Purples256\n\noutput_notebook()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train = pd.read_csv('/kaggle/input/bengaliai-cv19/train.csv')\ntrain.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"def plot_count_heatmap(feature1, feature2, train):  \n    count = train.groupby([feature1, feature2])['grapheme'].count().reset_index()\n    return count.pivot(feature1, feature2, \"grapheme\").fillna(0)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"def plot_heatmap(f1, f2, width, height, cbar=True):\n    f1_len, f2_len = len(train[f1].unique()), len(train[f2].unique())\n    \n    # index list\n    x = [str(i) for i in range(f2_len)] * f1_len\n    y = [str(i) for i in range(f1_len) for _ in range(f2_len)]\n    \n    # count list\n    tmp = plot_count_heatmap(f1,f2, train)\n    value = [tmp[int(a)][int(b)] for a, b in zip(x, y)]\n    \n    # example letter list\n    letter = train.groupby([f1, f2])['grapheme'].unique().unstack().fillna('')\n    lst = [','.join(letter[int(a)][int(b)]) for a, b in zip(x, y)]\n    \n    # processing for bokeh\n    df = pd.DataFrame({f2 : x, f1 : y, 'count' : value, 'example': lst})\n    source = ColumnDataSource(df)    \n    \n    # make continuous color palette\n    colors = list(reversed(Purples256))\n    mapper = LinearColorMapper(\n        palette= colors,\n        low=min(value),\n        high=max(value)\n    )\n\n    # make figure\n    p = figure(title=f\"{f1} & {f2} Count Heatmap\", tools=\"hover\", \n               toolbar_location=None,\n               x_range=list(map(str, tmp.columns)), y_range=list(map(str, tmp.index))[::-1],\n               plot_width=width, plot_height=height\n    )\n\n    # heatmap\n    p.rect(f2, f1, 0.95, 0.95, source=source,\n          fill_color={'field': 'count', 'transform': mapper}, \n           line_color=None)\n    \n    # tooltips\n    p.hover.tooltips = [\n        (\"Count\", \"@count\"),\n        (f1, f\"@{f1}\"),\n        (f2, f\"@{f2}\"),\n        (\"Example\", \"@example\")\n    ]\n\n    # detail setting\n    p.grid.grid_line_color = None\n    p.axis.axis_line_color = None\n    p.axis.major_tick_line_color = None\n    p.axis.major_label_text_font_size = \"5pt\"\n    p.axis.major_label_standoff = 0\n\n    # colorbar\n    \n    color_bar = ColorBar(color_mapper=mapper, major_label_text_font_size=\"5pt\",\n                         ticker=BasicTicker(desired_num_ticks=len(colors)),\n                         formatter=PrintfTickFormatter(format=\"%d\"),\n                         label_standoff=6, border_line_color=None, location=(0, 0))\n    if cbar : p.add_layout(color_bar, 'right')\n    \n    show(p)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plot_heatmap('grapheme_root', 'vowel_diacritic', 600, 3000)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plot_heatmap('grapheme_root', 'consonant_diacritic', 400, 3000)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plot_heatmap('vowel_diacritic', 'consonant_diacritic', 400, 500, False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":1}