by Olivia Jonah, April 2017
This dataset is about the information of baseball players: player's name, handedness (right, left handed, or both), height, weight, batting average, and the number of home runs.
The relation between batting average (avg) and home runs (HR).
Players with a hiigher avg tends to hit more home runs than lower avg players.
this shows that this belief that many great hitters were left-handed holds.
This visualization, shows that there is pretty high home run ratio of left-handed among top batting average players.
Data downloaded: baseball_data.csv the dataset recommended by Udacity.
It has Player's physical information(name, height, weight) and performance information(avg, HR).
| height | weight | avg | HR | |
|---|---|---|---|---|
| count | 1157.000000 | 1157.000000 | 1157.000000 | 1157.000000 |
| mean | 72.756266 | 184.513397 | 0.186793 | 45.359551 |
| std | 2.142272 | 15.445995 | 0.106175 | 74.065110 |
| min | 65.000000 | 140.000000 | 0.000000 | 0.000000 |
| 25% | 71.000000 | 175.000000 | 0.138000 | 1.000000 |
| 50% | 73.000000 | 185.000000 | 0.238000 | 15.000000 |
| 75% | 74.000000 | 195.000000 | 0.258000 | 55.000000 |
| max | 80.000000 | 245.000000 | 0.328000 | 563.000000 |
There are 1157 Players' records. Records with Players who have zero batting average or home runsare removed.
| height | weight | avg | HR | |
|---|---|---|---|---|
| count | 871.000000 | 871.000000 | 871.000000 | 871.000000 |
| mean | 72.338691 | 182.846154 | 0.244201 | 60.253731 |
| std | 1.935478 | 14.971904 | 0.030703 | 79.940631 |
| min | 65.000000 | 140.000000 | 0.104000 | 1.000000 |
| 25% | 71.000000 | 170.000000 | 0.232000 | 11.000000 |
| 50% | 72.000000 | 180.000000 | 0.248000 | 27.000000 |
| 75% | 74.000000 | 193.000000 | 0.263000 | 78.000000 |
| max | 80.000000 | 230.000000 | 0.328000 | 563.000000 |
After cleaning ie removing records with 0, there 871 records left.
Take a look at Player's performance with different handedness and avg_category.
A histogram shows, HR vs. avg_category with 3 types of handedness.
Using pandas.cut to split avg to 5 grades from 0.1 to 0.35.
| avg_category | handedness | HR |
|---|---|---|
| [0.1, 0.15) | B | 9 |
| L | 11 | |
| R | 41 | |
| [0.15, 0.2) | B | 2 |
| L | 32 | |
| R | 279 | |
| [0.2, 0.25) | B | 904 |
| L | 4449 | |
| R | 9020 | |
| [0.25, 0.3) | B | 2268 |
| L | 12591 | |
| R | 21861 | |
| [0.3, 0.35) | B | 160 |
| L | 660 | |
| R | 194 |
Since the population is unbalanced, the values need to be normalized,
by adding HR_ratio which is divided by subtotal HR sum for each handedness group.
| avg_category | handedness | HR | HR_partial sum | HR_ratio |
|---|---|---|---|---|
| [0.1, 0.15) | B | 9 | 61 | 0.147541 |
| [0.1, 0.15) | L | 11 | 61 | 0.180328 |
| [0.1, 0.15) | R | 41 | 61 | 0.672131 |
| [0.15, 0.2) | B | 2 | 313 | 0.006390 |
| [0.15, 0.2) | L | 32 | 313 | 0.102236 |
| [0.15, 0.2) | R | 279 | 313 | 0.891374 |
| [0.2, 0.25) | B | 904 | 14373 | 0.062896 |
| [0.2, 0.25) | L | 4449 | 14373 | 0.309539 |
| [0.2, 0.25) | R | 9020 | 14373 | 0.627566 |
| [0.25, 0.3) | B | 2268 | 36720 | 0.061765 |
| [0.25, 0.3) | L | 12591 | 36720 | 0.342892 |
| [0.25, 0.3) | R | 21861 | 36720 | 0.595343 |
| [0.3, 0.35) | B | 160 | 1014 | 0.157791 |
| [0.3, 0.35) | L | 660 | 1014 | 0.650888 |
| [0.3, 0.35) | R | 194 | 1014 | 0.191321 |
The first visualization graph has avg_category for x-axis, and HR_ratio for y-axis ande 3 groups visualization,
such as R, L, and B
avg_categoryis chosen as categorical variable, to simplify the visualization. Expressingavgas 5-graded levels is better than a value.- Bar chart gives a simplistic view by comparing certain quantity. one can easily spot out the highest value.
- Dimple.js is used to show the exact value of
HR_ratio, it gets hovering/tooltip function. - Different colors were used for
handedness, to be able to verify the ratio of handedness. Coloring is very effective way to show different groups data.
Family members and colleagues, appreicated what they saw
Your graph on the baseball record is interesting. But when I saw for the first time, I did not recognize at a glance what a graph is about. Perhaps, x-axis and y-axis label name seems to be important. In addition, if the short title could show about the contents of the graph, it would be great.
I recommend a pie chart showing the hit ratio. The pie chart is a very effective way to show the occupation state for some quantity.
The home run ratio with the handedness is impressive. I'm wondering one morething when I look this graph. The average number of home runs will be good index for figuring out difference between 3 handedness groups.
I made the follwing changes as reaction to feedbacks from my firends.
- I added short title to convey my discovery from this baseball dataset.
xandyaxes' labels are also included properly. - Also, I changed my plot to pie plot to show the home run ratio among
handedness. - From baseball dataset, I computed mean of homerun for each group, and set that variable to
y-axis.
This graph shows that there is a direct relation between batting average and Player's HRs. In addition, left-handed Players are especially noticed among top Players(high batting average).
Udacity/DataVis/baseball_data.csv: the original datasetUdacity/DataVis/cleaned_baseball.csv: grouped and merged dataset.