
import seaborn as sns
import pandas as pd
import matplotlib as plt
from matplotlib import style
%matplotlib inline


piece_code = "Q1"
move_color = "white"


df = pd.read_csv("C:\\Users\\chris\\Documents\\Datasets\\Chess PGN\\pgn_move_records.csv", dtype={"eco": str})
df = df[(df["piece_code"] == piece_code) & (df["move_color"] == move_color)]
df = df[["end_square.x", "end_square.y", "piece_code"]]
df.head()


df = df.groupby(["end_square.x", "end_square.y"]).count().reset_index()
df.rename(columns = {"piece_code" : "move_count"}, inplace = True)
df.head()


for x in range(1,9):
    for y in range(1,9):
        if (x not in list(df["end_square.x"])) or (y not in list(df["end_square.y"])):
            df2 = {"end_square.x" : x, "end_square.y" : y, "move_count": 0}
            df = df.append(df2, ignore_index = True)


df["end_square.x"] = df["end_square.x"].replace(1, "a")
df["end_square.x"] = df["end_square.x"].replace(2, "b")
df["end_square.x"] = df["end_square.x"].replace(3, "c")
df["end_square.x"] = df["end_square.x"].replace(4, "d")
df["end_square.x"] = df["end_square.x"].replace(5, "e")
df["end_square.x"] = df["end_square.x"].replace(6, "f")
df["end_square.x"] = df["end_square.x"].replace(7, "g")
df["end_square.x"] = df["end_square.x"].replace(8, "h")


df_pivot = df.pivot_table(index="end_square.y", columns="end_square.x", values="move_count", fill_value=0, dropna=False)
df_pivot


sns.set(rc = {"figure.figsize" : (12,12)})
sns.set(font_scale = 2)
heatmap = sns.heatmap(df_pivot, fmt="d", cmap="YlGnBu", square=True, xticklabels=True, yticklabels=True, linewidths=2, linecolor="black", cbar=False)
heatmap.invert_yaxis()
heatmap.set(xlabel=None, ylabel=None)
heatmap.set_title(f"Chess Piece - Placement Frequencies\n{piece_code} - {move_color.capitalize()}\n")
heatmap.tick_params(left=False, bottom=False)


fig = heatmap.get_figure()
fig.savefig("chess-heatmap.png")

end_square.x	a	b	c	d	e	f	g	h
end_square.y
1	4	6	7	36	14	7	3	3
2	2	6	30	86	134	18	14	4
3	4	36	46	112	64	156	38	31
4	46	31	52	119	60	48	73	22
5	16	45	30	68	52	35	34	74
6	24	14	43	28	36	35	20	28
7	33	59	31	27	29	53	44	32
8	14	10	11	40	18	13	9	19

Import Libraries.¶

Import dataset. Filter by piece and piece color. Drop excess columns, keeping only end square and piece code.¶

Group by square and count occurrences.¶

Fill our table with zeros and replace our x-values with letters so our heatmap resembles a chessboard.¶

Create a pivot table from our data. Fill empty/NA cells with 0.¶

Create heatmap using Seaborn.¶